JP3936833B2 - Body motion sensing device and body motion sensing system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention detects human body movement (body movement).Body movementSensingDevice and body motion sensingIt is related to the system, especially detecting the movement of a person's body according to the person's working condition and health condition.Body movementSensingDevice and body motion sensingAbout the system.
[0002]
[Prior art]
Detecting body movement in a person's daily life and production activities is effective, for example, for the prevention of adult diseases that are triggered by lack of exercise and the health management of the elderly. This is important for realizing data collection for improving work efficiency in the field. In recent years, various portable body motion sensing systems that are portable and small and multifunctional for the purpose of practical body motion detection indoors and outdoors have been researched and proposed.
[0003]
As a typical example of a conventional technique related to a body motion monitoring (synonymous with sensing) system, for example, it is described in Fig. 2 of Journal of Mechanical Engineering & Technology, 21 No.3-4 (1997) pp.96-105. Technology is known (hereinafter referred to as Conventional Technology 1). This prior art 1 monitors the output signals of an acceleration sensor provided at three places of a person's wrist, ankle, and waist and an electrocardiogram waveform (ECG) by wire through signal wiring, and carries a plurality of monitored information. This is a system in which data is stored in multiple channels on an easy small semiconductor storage medium. The portable storage device described above is driven by a lithium battery or the like, and the stored information is subjected to statistical analysis after being transferred to a separate computer. In this prior art 1, by providing acceleration sensors at different positions on the body, it is possible to correctly monitor not only the motion of the waist close to the center of gravity of the person but also the motion caused by the motion of the hands and feet.
[0004]
As another conventional technique, there is one that attempts to distinguish various walking states from the output of one sensor instead of providing different sensors in various parts of the body. As this type of conventional technology, for example, a technology described in Method of Information in Medicine, 36 (1997) pp. 356-359 is known (hereinafter referred to as Conventional Technology 2). In this prior art 2, an acceleration sensor having sensitivity in a uniaxial direction is provided only at the waist position, and the output signal waveform is processed to distinguish walking on a flat ground, walking up a stairs, and walking down a stairs. Is possible. And this prior art 2 can actually distinguish a walking state by using wavelet conversion. In this prior art, a sensor is connected to an amplifier through a signal wiring of 10 m. However, in comparison with the prior art 1, it is not necessary to wire a plurality of sensors in each part of the human body. Therefore, there is an advantage that the movement is not hindered.
[0005]
[Problem to be solved]
Although the above-described prior art 1 can detect detailed body movements including movements of limbs, it has been necessary to arrange a plurality of sensors by wire to each part of the human body in order to prevent a complicated movement of the human body. For this reason, there has been a problem that it is not suitable for use to continuously detect natural body movements in daily life or production activities of individuals.
[0006]
Further, in the above-described conventional technique 2, since the output from one sensor is signal-processed to obtain a plurality of body movement information, there is a possibility that the sensor can be mounted without hindering an individual's daily life or production activity. There is. However, this prior art 2 focuses on the hip movement close to the center of gravity of the person and extracts and detects a signal from an acceleration sensor sensitive only to one axis, so it cannot capture various movements of each part of the human body. Has a problem.
[0007]
That is, it is difficult for both of the above-described conventional techniques to simultaneously detect the movements of various parts of the human body and to reduce the number of detection units (acceleration sensors, etc.) required for this purpose. When considering the practical use of a health management system or work management system, it has the problem that it is difficult to obtain the required body movement information of each part of the human body without affecting the body movement of the subject. ing.
[0008]
In order to construct a body motion sensing system that attaches a sensor to only one part of the human body and processes the signals while capturing the body motion with the sensor, it can grasp the movement of each part of the human body. When considering whether the problem should be solved, the following two problems can be considered.
[0009]
(1) The human center-of-gravity movement gives substantially the same periodic fluctuation signal regardless of the position of the detection unit, except for the magnitude of detection sensitivity and the presence or absence of disturbance. Since the position of the waist close to the center of gravity is appropriate in terms of sensitivity, the sensor is often mounted at the position of the waist. On the other hand, the movement of each part of the human body other than the center of gravity may change the characteristics of the signal waveform itself depending on the position of the detection unit. For example, when lifting an arm, the acceleration sensor attached to the wrist directly detects the change in acceleration accompanying the direct movement of the arm, but at the same time, the acceleration sensor attached to the back shows the change accompanying the movement of the muscles of the back. To detect. When considering the portability of the body motion sensing system, a small detector can be mounted not only on the waist belt, but also on the chest pocket, for example. The degree of freedom of implementation is required. However, in the case of trying to analyze various body movements by extracting the characteristics of the signal waveform as in the wavelet transform described in the prior art described above, the difference in the mounting position of the detection unit is the characteristic of the signal waveform. Because it causes a change in itself, it has a great influence on detection results such as false detection.
[0010]
(2) When trying to capture the movement of each part of the human body other than the center of gravity, there are cases where various body movements cannot be clearly captured only by sensor output having sensitivity only in one axial direction. When the center of gravity movement during walking is the main measurement object as in the prior art described above, the direction of vibration of the body can be represented by the movement of almost one axial direction, so the direction in which the sensor has sensitivity is the direction. And the output signal may be processed. On the other hand, in order to sense more complex body movements, it is generally necessary to obtain outputs in three independent axial directions. In this case, for example, a method of carrying a plurality of independent sensors together in one place on mounting is conceivable. However, if the wavelet transform described above is applied to each sensor output as it is, the load of signal processing increases. . Further, when considering a maintenance-free body motion sensing system for a long period of time, it is necessary to consider a power supply method sufficient for driving the plurality of detection units described above. In addition, in order to consider a device that is convenient to carry, a detection unit for obtaining information needs to be able to detect a plurality of pieces of information and be mounted in a small size.
[0011]
  The purpose of the present invention is toBody motion sensing device and body motion sensing capable of grasping the motion of a human body even when the position where the device for identifying body motion is mounted is not always constant and changesTo provide a system.
[0012]
[Means for Solving the Problems]
  According to the present invention, the object is a portable body motion sensing device having at least one information detection unit and a signal processing unit that processes an output signal of the information detection unit to obtain body motion information.Mounting position input means for inputting the mounting position of the information detection means;The signal processing means includesFrom the mounting position input meansBased on the mounting position information of the information detecting means,Has body motion recognition means to recognizeIs achieved.
[0013]
  Also, the purpose isIn a portable body motion sensing device having at least one information detection means and a signal processing means for obtaining body motion information by processing an output signal of the information detection means, an implementation for inputting a mounting position of the information detection means A body input recognition unit, a body movement recognition unit for recognizing body movement based on mounting position information of the information detection unit from the mounting position input unit, and a database corresponding to the mounting position; The means recognizes body movement based on the correlation between the database and the information from the information detection means.Is achieved.
[0014]
  Also, the purpose is information detectionmeansAnd the information detectionmeansProcessing to obtain body movement information by processing the output signalmeansIn the body motion sensing system, the information detectionmeansPosition information input that gives mounting position informationmeansThe signal processingmeansIs the location information inputmeansThis is achieved by estimating the motion of the human body part not mounted with the information detection means by performing signal processing in accordance with the position information from.
[0017]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of a body motion sensing system according to the present invention will be described in detail with reference to the drawings.
[0018]
FIG. 1 is a block diagram showing a configuration of a body motion sensing system according to a first embodiment of the present invention. In FIG. 1, 1 is information detection means, 2 is position information input means, 3 is signal processing means, 4 is display / communication means, 5 is a power supply unit, 7 is a portable device, 21 is a position information input unit, and 22 is a position. An information input unit 23 is a storage means.
[0019]
The body motion sensing system according to the first embodiment of the present invention is configured as a portable device 7 attached to a human body, and information detection means 1 for detecting body motion information such as human body displacement, velocity, acceleration, rotation angle, rotation angular velocity and the like. And position information input means 2 for inputting mounting position information of the information detection means 1, signal processing means 3, and external input / output means, for example, display on a liquid crystal screen, touch input on the same screen, wired wireless The display / communication means 4 that is an interface of the communication means and the power supply unit 5 are configured.
[0020]
In the above description, when using a plurality of detection elements (sensors) as described later, the information detection means 1 is mounted on almost one place. The position information input unit 2 includes a position information input unit 21 of the position information input unit and a position information control unit 22 that determines how to transmit the signal obtained from the position information input unit 21 to the signal processing unit 3. It is configured. The position information control unit 22 includes a storage unit 23 that can store the input position information in a nonvolatile manner. The signal processing means 3 refers to the position information 102 and 106 obtained from the position information input means 2 and the signal 104 input from the outside via the display / communication means 4 and outputs the output signal 101 of the information detection means 1. And outputs the result as a signal 103. The display / communication means 4 may include an I / O port in addition to the functions described above. When the portable device 7 is used without performing communication with the outside, the display / communication unit 4 may include a storage unit as a temporary output destination.
[0021]
Since the contents of the position information 102 may be switched during system operation, the position information control unit 22 always stores and holds new position information in the storage unit 23. Further, when the stored content is updated, the position information control unit 22 notifies the signal processing means 3 by the signal 107 that the stored content has been updated. As a method for transmitting the update of the stored content, for example, a flag switching process may be performed. It is also possible to transmit the updated information as it is to the signal processing means 3 as the signal 107. A signal 105 is an instruction for reading the stored position information, and a signal 106 indicates the read position information. Specifically, the power supply part 5 is comprised with the battery which can be charged / discharged. 1 indicates a power supply signal when power is supplied wirelessly from the outside of the apparatus via the display / communication means 4.
[0022]
Each component described above is integrally mounted as a portable device 7 so as to be carried at substantially the same position. The position information input unit 21 is simply a changeover switch provided in the portable device 7. For example, portable locations of portable devices are roughly classified into breast pockets, waist side surfaces, back surfaces of the waist, etc., and these are assigned to numbers 1, 2, and 3 shown in the figure, and the user switches between locations where they are actually used. In FIG. 1, the apparatus is turned off at the switch OFF position so that the user can simultaneously turn on the switch and select a mounting location. If the switching of the position information input means 21 is appropriate, the portable device 7 shown in FIG. 1 can detect, judge, and output more correct body motion information for various forms of body motion.
[0023]
For example, in the case of body movements that are difficult to reflect in the center of gravity movement, such as shaking an arm, the detected body movement characteristics may be significantly different depending on the position of the mobile device 7 including the information detection unit 1. In such a case, the means shown in FIG. 1 can perform optimum signal processing according to the position of the apparatus by switching the position information input means 2. When switching is not appropriate, detailed body motion information cannot be obtained, but, for example, motion related to the center of gravity of the body such as periodic vibration accompanying walking can be detected almost accurately. The type of switching is not necessarily the type shown in the figure, and a finer section or a larger section may be provided. For example, you may divide roughly into two types, when attaching to a body and carrying in a bag. As explained below, if a signal processing method that analyzes and judges the characteristics of the waveform is adopted, it is possible to detect the required signal characteristics without using the body movement directly. It is possible to estimate a highly likely body movement.
[0024]
  As described above, the system according to the embodiment of the present invention can obtain an appropriate output result for the motion near the center of gravity of a person regardless of whether the switching position of the position information input unit 21 is correct or not. For this reason, for example, pay attention to periodic body movement changesStepsA user who only needs relatively rough information as in the case of using the meter may move the switch with the same feeling as a normal ON switch without minding the switching position. In addition, a user who wants more detailed information about calorie consumption and the like can obtain more accurate information by frequently switching the switch, and thus can meet a wide range of user requests.
[0025]
Of course, the embodiment of the position information input unit 21 is not limited to the method shown in FIG. For example, the position information input unit 21 may be provided with a plurality of switches corresponding to the switching positions so that the user selects and presses them, or the number of times one switch is pressed OFF → 1 → 2 → 3 → OFF may be selected. Alternatively, the user may select a menu by touching the screen from a menu displayed on the screen. Further, unlike the case shown in FIG. 1, the ON / OFF switch of the portable device 7 may be provided separately from the position information input unit 21. In this case, as an example, two buttons, ie, an ON / OFF switching button (switch) and a button (switch) for repeatedly switching position information from 1 → 2 → 3 → 1 may be provided. When the ON / OFF switch of the portable device 7 is provided separately from the position information input unit 21, it is not necessary to input the position information every time it is used. Following the switch ON, if there is no new input 102 from the position information input unit 21, the storage means 23 already stores the position information up to the previous time. If the necessary read instruction 105 is given to the position information 106, the previously stored position information 106 can be obtained. In this way, it is only necessary to input the position information only for the first time, such as when the user customarily holds the portable device at substantially the same position. When there is a new input 102 in the middle of signal processing, the position information control unit 22 updates the contents of the storage means 23 and transmits the update of the stored contents to the signal processing means 3 with a signal 107.
[0026]
The signal processing means 3 may receive the update of the stored contents and read out new position information after a series of signal processing. The stored position information may be displayed on the display unit 4 as it is. The user can carry the device with confirmation and consent. The switching-type position information input unit 21 shown in FIG. 1 does not necessarily require the storage means 23 because the position information is mechanically held depending on the position of the changeover switch. In this case, when the switch is turned off once and then turned on again, the position information is naturally re-input according to the position of the switch simultaneously with the operation of turning on the switch as described above. The signal processing means 3 performs signal processing with reference to the position of the switch.
[0027]
The power supply unit 5 is supplied with power by the power supply signal 201 from the display / communication means 4 because the display / communication means 4 basically receives a signal transmitted from the outside in the form of an electromagnetic wave. Can be realized. In this case, the display / communication means 4 receives a sinusoidal electromagnetic wave having a specific frequency transmitted from the outside, obtains a sinusoidal voltage signal via a sinusoidal current or resistance, and rectifies this to generate a direct current. A voltage can be obtained and used for charging the battery of the power supply unit 5 that can be charged and discharged. Although the power that can be transmitted wirelessly is generally small, it can be used for long-distance communication. For example, it can be used as a power supply means for a system that combines a battery-centric CMOS-based signal processing circuit that can save power. Can do. The power supply may be performed at all times, but the remaining charge of the power supply unit 5 is monitored by a voltage value or the like, and as a result, a request to start charging is made when the remaining charge becomes a predetermined value or less. It may be transmitted to a predetermined external device.
[0028]
And, since the situation of the subject can be determined from the body movement information detected by the illustrated system, depending on the result, wireless power supply may or may not be performed. it can. Since the power supply by wireless transmits a certain amount of power, it can be assumed that the wireless affects a precision machine or an electric appliance. For this reason, for example, when vibration of a train is detected from body movement information, it can be determined that the target is on the train, and when it is determined that the subject is under a predetermined condition, The communication mode can be switched, for example, by not transmitting the request to start charging as described above. Here, the communication mode is defined by classifying conditions for communication with the outside such as limiting communication exceeding a predetermined power. A condition relating to the content of information such as communicating only predetermined information may be set as the communication mode. The signal 103 output during a period of no communication with the outside may be stored in a storage unit that can be provided as a temporary output destination in the display / communication unit 4. The information temporarily stored in the display / communication means 4 may be communicated collectively when it is determined that communication with the outside is possible, or the information may be stored in the stored state. You may make it utilize as.
[0029]
In the embodiment shown in FIG. 1, it is assumed that the output signal 103 is transmitted to the outside continuously or at predetermined intervals or in response to an external request signal. You may make it preserve | save in the memory | storage means in the above-mentioned display and communication means 4 which is an output destination. The condition for determining which communication mode is selected in what situation is determined in advance in the signal processing means 3 or the display / communication means 4.
[0030]
FIG. 2 is a diagram for explaining a configuration example of the information detection means 1 provided in the first embodiment of the present invention, which will be described below. In FIG. 2, 60 is a case, 61a is a pressure detection means, 61b is a diaphragm, 62 is a silicon substrate, 63 is a signal conversion means, 64 is a mass body, 65 is a deformed body, 66 connection means, and 67 is a connector. 2A shows a cross-sectional view of the information detecting means, and FIG. 2B shows a top view thereof.
[0031]
When detecting complex body movements, it is preferable that a plurality of pieces of independent information such as vibrations and rotational movements with three degrees of freedom can be detected simultaneously. On the other hand, if a large number of sensors are collected and mounted as the information detection means 1 for that purpose, the apparatus may be increased in size and a problem of cost increase may occur. The information detection unit 1 shown in FIG. 2 is configured so that vibration changes with three degrees of freedom of x, y, and z can be collectively captured by one sensor chip. The information detecting means 1 includes a dome-shaped pressure detecting means 61a provided on a silicon substrate 62 housed in the case 60, and a circular diaphragm 61b provided on the dome-shaped ceiling of the pressure detecting means 61a. And a signal conversion means 63 provided on the silicon substrate 62, a mass body 64, and a deformation body 65.
[0032]
In the above description, the structure of the diaphragm 61b is not necessarily circular, and may be, for example, a square or a general polygon. The diaphragm 61b is a pressure detection surface. Since the diaphragm 61b is deformed when it receives pressure, the pressure change can be electrically detected by detecting the deformation through the piezoresistance effect or the capacitance change. The structure of the pressure detection means 61a can be formed on the surface of the silicon substrate 62 by using a surface micromachining technique called sacrificial layer etching. The signal conversion means 63 provided on the silicon substrate 62 converts the change in capacitance (C) into the change in voltage (V) when the pressure detection means 61a detects the deformation of the diaphragm 61b by the change in capacitance. Since this is convenient for handling signals, the conversion is performed. Further, the signal conversion means 63 compensates for the linearity of the sensor output since the sensor output as the pressure detection means 61a generally shows a non-linear response, and adjusts the slope and zero point of the linear output after compensation. Process. The signal conversion means 63 is obtained by integrating functions of an A / D conversion unit, a memory unit, an arithmetic circuit, a D / A conversion unit, an analog circuit, a protection circuit, and the like for performing these processes by a semiconductor process. Of course, a more advanced signal processing circuit may be added to the signal conversion circuit 63.
[0033]
The mass body 64 is provided so as to be in contact with the deformable body 65 covering the diaphragm 61b on the pressure detecting means 61a described above, and deforms the deformable body 65 in accordance with the body movement. The mass body 64 is formed of, for example, a metal plate such as aluminum, and the deformable body 65 is, for example, a silicone gel. In view of environmental resistance, a fluorine-based silicone gel is desirable. The mass body 64 and the deformable body 65 may be mounted so as to be in direct contact with each other as shown in FIG. 2A, or a thin film is provided on the surface of the deformable body 65, and the mass is formed on the thin film. A body 64 may be provided. Alternatively, the mass body 64 may be provided inside the deformable body 65.
[0034]
The silicon substrate 22 in the illustrated information detecting means 1 and the outside thereof are electrically connected to a connector 67 for connection with the outside by a connecting means 66. The connection means 66 is, for example, a gold wire or an aluminum wire, and the connection can be realized by wire bonding. Although not shown in FIG. 2, it is desirable to provide a lid on the case 60 so that the mass body 64 and the deformable body 65 are not touched by a finger. Further, the pressure detection means 61a and the signal conversion circuit 63 described above can be integrated on a small silicon substrate 62 having a size of 3 mm square and a thickness of 0.5 mm, for example. 7 can be made sufficiently small.
[0035]
Next, the function of the information detection means shown in FIG. 2 will be described. Now, for example, if there is a body movement in the x (or y) axis direction in FIG. 2B, the mass body 64 has inertia, so the body 65 moves opposite to the body movement while pulling the deformable body 65. , Give distortion inside the deformation body. Similarly, if there is a body motion above the z axis perpendicular to the paper surface of FIG. 2B, the mass body 64 compresses the deformable body 65, and if there is a body motion below the z axis, the mass body 64 is deformed. It moves relative to pull the body 65. The deformation of the deformable body 65 causes deformation of the diaphragm 61b of the pressure detecting means 61a through an internal stress change. Therefore, if this is electrically detected, the occurrence of body movement can be detected. In general, since the body motion changes applied to the x, y, and z axes are superimposed, the pressure detection unit 61a outputs a complex signal waveform, but the waveform of the signal output along with the characteristic body motion is used as a reference. If stored in advance as a waveform, a specific body motion can be extracted and separated by comparing the correlation between the reference waveform and the signal waveform. According to the first embodiment of the present invention, a change in body motion corresponding to three spatial degrees of freedom in one output waveform of the sensor is strengthened in a certain part and weakened in a certain part to become a specific signal. By extracting the characteristics of the signal, it is possible to comprehensively grasp the movement of the whole body.
[0036]
Here, consider the same motion in the x-axis direction and the y-axis direction. If the center of the mass body 64 and the center of the diaphragm 61b coincide with each other in the xy plane, in this case, the deformation body 65 is subjected to the same deformation that differs only in the direction. Can not do it. In order to avoid this, a diaphragm having no axial symmetry may be designed, or the hardness and type of the deformable body 65 may be made non-axisymmetric, but for simplicity, FIG. As shown, the centers of the mass body 64 and the diaphragm 61b are shifted, and in particular, as shown in FIG. 2 (b), the dimension a in the y direction between the mass body 64 and the center of the diaphragm 61b and the x direction What is necessary is just to make the dimension b different. Thereby, the degeneracy of sensitivity in the x-axis direction and the y-axis direction can be solved. More generally, an axis that passes through the center or center of gravity of the mass body 64 and is orthogonal to the diaphragm 61b that is the pressure detection surface of the pressure detection means 61a, and the center of the diaphragm 61b that is the pressure detection surface of the pressure detection means 61a. Alternatively, the axis passing through the center of gravity and orthogonal to the pressure detection surface may be prevented from overlapping each other.
[0037]
The normal usage of the sensor is to match the sensitivity of the x-axis and the sensitivity of the y-axis as accurately as possible and to process both signals independently. There is no need to calibrate the sensitivity. This is because it is more important to obtain a characteristic signal for each body movement than to match the sensitivity if the correlation evaluation between a reference waveform prepared in advance and a signal waveform is assumed. By the method of signal superposition, waveform synthesis can be performed so as to rather emphasize the characteristic change of a specific signal. Such a method is possible because the body motion to be noticed can be relatively limited for each system application, and it is not always necessary to detect all body motions uniformly.
[0038]
When it is necessary to capture changes in each axial direction independently, a plurality of pressure detecting means 61a may be provided on the silicon substrate, and these signals may be taken out independently. Since the sensitivity differs depending on which part of the silicon substrate is provided with the pressure detection means, the motion of each of the three degrees of freedom can be extracted by comparing the obtained signals.
[0039]
When a capacitance type pressure measurement is used as the pressure detection means 61a, a high detection sensitivity can be obtained in principle with a dome structure having a low step. This is because if the electrodes opposite to each of the diaphragm 61b and the substrate 62 are provided to capture the change in capacitance, the detection sensitivity to deformation can be increased as the distance between the electrodes is narrowed. In order to accurately grasp the deformation of the deformable body 65, a diaphragm having a large aperture and a low step may be provided. Since the frequency responsiveness of the diaphragm itself can be made sufficiently high, the frequency responsiveness as the pressure detecting means can be adjusted by the amount and hardness of the silicone gel that is the deformable body 65.
[0040]
FIG. 3 is a diagram for explaining an example of signal processing in the first embodiment of the present invention, which will be described below. In FIG. 3, 31 is an A / D conversion processing unit, 32 is a feature amount extraction processing unit, 33 is a body movement determination processing unit, 34 is a database selection processing unit, 35 is a database set, and other symbols are shown in FIG. It is the same as the case.
[0041]
As shown in FIG. 3, the signal processing unit 3 described above includes an A / D conversion processing unit 31 that performs A / D conversion on the signal from the information detection unit 1, and a signal from the A / D converted signal. A feature amount extraction processing unit 32 that extracts feature amounts, a body motion determination processing unit 33 that determines body motion based on the feature amount of the extracted signal, a database selection processing unit 34, and a database set 35 are included. It is configured. At least one of the feature quantity extraction processing unit 32 and the body motion determination processing unit 33 performs signal processing by referring to at least one database included in the database set 35 stored in advance (108, 109). For example, when performing the correlation evaluation with the reference data, the feature amount extraction processing unit 32 switches the reference data according to the position information. In addition, the body motion determination processing unit 33 selects and switches the type or threshold value of the parameter according to the position information when determining the body motion by comparing the value of a specific parameter with a threshold value. Here, the database to be referred to is selected from a plurality of database sets by the database selection processing unit 34 based on the mounting position information 102 obtained by the position information input means 2. The number of database sets may be set according to the required type and accuracy of body motion detection. The database selection processing unit 34 refers to the mounting position information stored in the storage unit 23 included in the position information control unit 22 when selecting a database. When a new signal is input to the position information input unit 2, the position information control unit 22 outputs a signal 107 indicating that the position information has been changed. Upon receiving this signal 107, the database selection processing unit 34 refers to the storage means 23 (105) and obtains position information (106).
[0042]
The first embodiment of the present invention employs a method for determining body movement by comparing features of signal data stored in advance and signal data obtained by measurement. The feature is that the database is selected by switching the database set 35 based on the mounting position information obtained from the position information input means 2. In general, feature quantities such as vibration components associated with body movement vary depending on the position of the portable body motion sensing system that is carried. Can be determined correctly.
[0043]
The above-described signal processing process is premised on digital signal processing, but the same processing may be performed centering on an analog circuit. For example, when a signal generating means for generating a reference signal and a lock-in amplifier that outputs a correlation strength by calculating a product of the reference signal and the input signal are combined, a correlation evaluation between the reference signal and the input signal is performed. be able to. In this case, the reference signal generated by the signal generation means may be switched based on the input signal 102 of the position information input means 2. Further, instead of switching the reference signal generated by one signal generating means, a plurality of signal generating means may be provided so that the reference signals output from the respective signal generating means are appropriately switched and used.
[0044]
When the body motion determination processing unit 33 performs body motion determination, the determination may be facilitated by the reference information 104 from the outside. For example, in the work management system, if the worker is in which compartment, the work content for each compartment may be limited, and if the work in a specific compartment is only opening and closing of a valve, Periodic arm movements indicate that the valve is being opened and closed. In some cases, the open / close direction can be estimated from the dominant arm and its movement. When such reference information 104 is given from the outside, the body motion determination processing unit 33 can perform detailed estimation of the body motion.
[0045]
FIG. 4 is a diagram for explaining an example of correlation evaluation in the frequency domain, which will be described below.
[0046]
In the example shown in FIG. 4, the spectrum intensity information of the reference signal is used as the database. Specifically, as shown as reference data in FIG. 4A, a set of databases in which signal intensities are assigned to each frequency region separated from region 0 to n (integer) is used as one database. Here, the spectrum intensity F (m) in the frequency region m is normalized so that the sum of the regions is 1. This database represents, for example, one body movement mode such as walking. In the example shown in the figure, the database is characterized in that the intensity of the frequency domain k is significantly larger than the surrounding frequency domain. In order to perform comparison processing with reference data included in the database, the signal obtained from the information detection unit 1 is also converted into a frequency domain signal. Specifically, after the A / D conversion by the A / D conversion unit 31 described with reference to FIG. 3, the feature amount extraction unit 32 obtains a spectrum intensity distribution using a fast Fourier transform (FFT) technique. To do. An example of the result is schematically shown as observation data in FIG. According to this observation data, it can be seen that the signal intensity in the frequency domain k is as small as the value in the surrounding frequency domain, so that it is different from the body motion mode corresponding to the reference data.
[0047]
The comparison between the reference data and the observation data is performed by the body movement determination unit 33 shown in FIG. At this time, for example, if the information 102 “back of hip position” is input from the position information input means 2, the information is stored in the storage means 23, so the database selection unit 34 Choose the right database for your information. Here, the database suitable for the position information is, for example, that the body motion sensing system according to the present invention is previously attached to the “back of hip position” and the detection signal 101 is collected, or the same conditions are reproduced by simulation. It means a database created by predicting the detection signal 101. Instead of using the obtained information as it is as a database, only the signal features may be extracted and processed into a database. The selected database is used for body movement determination by the body movement determination unit 33 (109). In this case, the body movement determination is to compare the spectrum intensity distributions F (m) and G (m) between the reference data and the observation data.
[0048]
Specifically, as shown in FIG. 4 (a) and FIG. 4 (b) as a calculation method, the smaller one of the signal strengths F (m) and G (m) is selected for each frequency region m. Then, as shown in FIG.4 (c), both overlap part H (m) can be extracted. The greater the overlap, the greater the correlation. Therefore, the value obtained by calculating the sum of H (m) for the frequency domain m can be used as an index of correlation strength. In the case of the example shown in FIG. 4, since the signal intensity in the frequency domain k is small in the observation data as already described, the correlation intensity is small. The body motion determination by the body motion determination unit 33 is observed with the body motion mode having the greatest correlation among the representative body motion modes included in the selected database by repeating the comparison process as described above. It is performed so that it may be judged as body movement. In this case, it is preferable that the correlation strength index is larger than a certain threshold. This is because the actual body movement may not match any body movement mode prepared in advance. In such a case, the obtained indexes are all small, but have some value due to the influence of signal noise or the like, and if they are compared with each other as a reliable value, an incorrect result is output. If the condition is that the index is larger than a certain value, only a significant signal can be used for the determination process.
[0049]
The correlation evaluation described above may be performed only for a specific frequency region. In the above example, a specific body motion mode can be extracted by paying attention to the frequency domain k. One frequency region or a plurality of frequency regions to be noted can be selected with reference to information (102, 107, 109) obtained from the position information input means 2 in body motion determination.
[0050]
In the above-described embodiment of the present invention, when the actual body motion does not match any of the body motion modes prepared in advance, the information “distinguishable” is output. On the contrary, the result of indistinguishability includes information that “does not match any body motion mode prepared in advance”, so it can be effectively used as one of observation information even in the case of system application described later. Can do.
[0051]
The number of body motion modes prepared for each of the plurality of databases can be determined more accurately as the number increases, but there is a problem that the amount of the database becomes enormous. In this case, the amount of data can be reduced by selecting a representative body motion mode according to the purpose of the system. Although limiting the body motion mode seems to reduce the body motion to be discriminated, the actual body motion can often be expressed by a combination of basic modes. In such a case, more complex operation can be discriminated by considering not only the maximum value of the correlation strength index but also the series of modes and the ratio of the index value to each mode. For example, if two body movement modes occur, such as bending down and moving the hand, and the index value has a predetermined size and ratio, this is recognized as one action. That's fine. If the work environment information can be obtained as the reference signal 104, it is possible to distinguish whether this operation is a “grasping grass” operation or an operation of “opening / closing the foot piping valve”. A series of movements can be represented by changing the sum of the indices or the weighted sum for each mode into one index, and a combination of basic body movement modes can be represented by one action. By recognizing as, it is possible to express advanced body movements without greatly increasing the amount of the database.
[0052]
FIG. 5 is a diagram for explaining an example of correlation evaluation in the time domain, which will be described below. In FIG. 5, 32a is a correlation evaluation part, 32b is a bias detection part, and another code | symbol is the same as the case of FIG.
[0053]
In the example shown in FIG. 5, a plurality of reference signal waveforms having a finite length are used as a database. These reference signal waveforms are obtained by extracting and processing the characteristics of the signal waveforms obtained for each body motion mode depending on the carrying position of the body motion sensing system through experiments or simulations. The signal 106a is mounting position information for selecting a reference waveform or a set of reference waveforms. The signal 106 b is mounting position information referred to in the body movement determination unit 33. The signals 106a and 106b are provided from the position information control unit 22 already described to the reference waveform selection unit 34 and the body motion determination unit 33, respectively. In FIG. 5, a signal 105 related to information reading from the position information control unit 22 is not shown. The correlation evaluation unit 32a performs correlation evaluation between the detection signal waveform and the reference waveform. The body movement determination unit 33 changes the determination threshold based on the mounting position information 106b, for example. Depending on the mounting position of the body motion sensing system, the detection signal of a specific body motion mode may be relatively small. However, if an appropriate threshold is set with reference to the mounting position information 106b, more correct body motion detection is possible. Can do. The contents that can be determined with reference to the mounting position information 106b are not limited to the threshold setting. For example, when the correlation evaluation unit 32a performs correlation evaluation with a plurality of reference waveforms, the weight for each reference waveform can be changed depending on the mounting position information 106b. This makes it possible to obtain a correct evaluation result even when the characteristics of the signal waveform obtained depend on the portable position of the system.
[0054]
In the evaluation by the correlation evaluation unit 32a, the direct current (bias) component of the detection signal is detected and removed in advance by the bias detection unit 32b, so that only the AC characteristic of the signal can be easily extracted. . That is, as the reference signal waveform, only the AC characteristics of the signal need be extracted. On the other hand, the DC (bias) component of the signal often contains useful information. For example, when an acceleration sensor is used for signal detection, it is possible to determine whether a person is standing or lying from the value of the direct current (bias) component by devising the mounting direction of the sensor. The body movement determination unit 33 uses the value of the direct current (bias) component of the signal separated by the bias detection unit 32b in the determination process together with the correlation evaluation. Specifically, even if the alternating current component is almost zero and the correlation evaluation result is indistinguishable, if a significant value is shown in the direct current (bias) component, for example, a value corresponding to gravitational acceleration, It can be seen that the acceleration sensor is in a direction in which gravity can be detected, and as a result, for example, it can be determined that a person is lying down. In addition, if there is a gradual change corresponding to the breathing cycle or a change such as turning over in the alternating current component, it is understood that the person is sleeping comprehensively. On the other hand, if the AC component is almost zero (a predetermined threshold or less) for a certain time or more, it can be estimated that an emergency state such as an accident has occurred.
[0055]
There are various correlation evaluation methods other than those described above. Next, a method by direct comparison between a reference waveform and an observation signal waveform and a method by wavelet transform will be described.
[0056]
The method of directly comparing the reference waveform and the observation signal waveform evaluates the magnitude of correlation between a waveform sampled for a finite time and each reference waveform prepared in advance. For the correlation evaluation method, for example, the vertical axis of the data shown in FIGS. 4 (a) and 4 (b) is changed from the spectrum intensity to the signal intensity for each time, and the horizontal axis is discretized from the discretized frequency. The method described with reference to FIG. 4 can be used as it is. The problem here is that the time required for each body movement mode is different. As a method for efficiently detecting body motion modes of different required times, there is a method of cutting out an observed waveform for the required time, performing waveform expansion / contraction, that is, time normalization processing, and normalizing the signal waveform with respect to the required time. The required time for each body movement mode is, for example, the time when the differential value of the signal after averaging the noise has reached a predetermined threshold or more from the state of almost zero is set as the start of the time required for the body movement mode, and is zero again. The time required to return to the state can be set as the required time. In the case of a motion having an obvious periodicity such as walking, or a motion having a pseudo periodicity in which a repetitive object is lifted on a shelf, the required time may be determined by the unit of repetition.
[0057]
The repetition unit can be obtained, for example, by performing FFT on the sampling signal and obtaining it from a frequency domain value having a spectrum intensity protruding in a region away from the DC component, or by detecting the period using an autocorrelation function. You may do it. On the contrary, for a signal that does not have a prominent peak in spectrum intensity, or for a signal that has a peak concentrated only in the vicinity of the DC component, attention should be paid to the change in the signal differential component as described above. The method of directly comparing the reference waveform and the observation signal waveform is suitable for detailed analysis of a signal having no remarkable periodicity.
[0058]
The wavelet transform method can be executed by selecting a basic wavelet type based on the signal 106a. Alternatively, the determination area to be noted may be selected by the signal 106b. In order to make these contents easy to understand, the wavelet conversion will be briefly described. In the wavelet transform, first, a wave packet shape that shows vibration for a finite time is selected. This is called a wavelet. While moving the selected wavelet along the time axis, the correlation with the actual signal is evaluated. This is the basis of wavelet transformation. In the wavelet transform, the wavelet is further expanded and contracted (scale conversion) along the time axis, and the correlation evaluation between the expanded and contracted wavelet and the actual signal is repeated. As a result, the output result of the wavelet transform generally becomes a two-dimensional map. For example, taking time on the horizontal axis and scale conversion on the vertical axis, you can obtain information such as the occurrence of finer or coarser scale vibrations at a certain time centered on the basic wavelet. Can do. In actual wavelet selection, although a condition to be satisfied may be given depending on the convenience of calculation, basically, a waveform showing a strong correlation with the characteristics of a signal waveform to be measured may be selected.
[0059]
In the embodiment of the present invention, when the wavelet transform method is used, the wavelet is selected based on the signal 106a. Therefore, the characteristics of the signal waveform that differs depending on the mounting position of the body motion sensing system are described in detail. Can be caught. Further, since it is possible to select a characteristic scale and time of a signal generated for each notable body motion mode based on the signal 106b, it is difficult to be affected by noise mixed in the scale and time that are not related to the scale or time. Features can be obtained.
[0060]
FIG. 6 is a diagram for explaining an example of signal waveforms characteristic of the body movement mode. The example shown in FIG. 6 is an example of a signal when a uniaxial acceleration sensor is used as the information detection means 1, and the sensor has a significant sensitivity centering on a direction perpendicular to the ground. To do.
[0061]
The example shown in FIG. 6A shows an example when a person stands up from the chair, and the example shown in FIG. 6B shows an example when a person sits on the chair. When a person stands up from a chair, a signal waveform is obtained in which the first speed change, a short uniform motion, and the last speed change are combined. When a person sits on a chair, the direction of speed change is reversed, and a small impact vibration waveform accompanying sitting is obtained. In order to extract the body motion mode of standing from a chair, for example, the waveform shown in FIG. Actually, the amplitude of the obtained waveform is not necessarily constant, and a time scale change that sits fast and sits slowly is expected, but the wavelet transform enables temporal correlation evaluation including scale transformation. Therefore, it is possible to grasp at which point in time the same body movement mode has occurred without being affected by the speed of movement. Similarly, in order to extract the body movement mode of sitting on a chair, basically, the same correlation can be obtained by using a wavelet whose polarity is opposite to that of the basic wavelet. In addition, vibrations associated with seating appear on a fine scale in which the basic wavelet is compressed on the time axis. Therefore, if the correlations of different areas of the scale are taken together, the distinction from similar body movement modes can be made clearer. . Specifically, there is a significant correlation corresponding to the action of “sitting and squatting” on a large scale, and there is a significant correlation corresponding to “sitting vibration” even at a fine scale at the time of the correlation (particularly in the latter half). For example, a mode of “sitting on a chair” can be specified from possible body movement modes. In the example shown in FIG. 6, it is assumed that a uniaxial acceleration sensor is used for simplicity, but the same method can be applied to a more complicated signal waveform.
[0062]
FIG. 7 is a block diagram showing a configuration of a body motion sensing system according to the second embodiment of the present invention. 7, 1a, 1b, 1c,... Are information detecting means, 11 is a signal synthesizing means, and the other symbols are the same as those in FIG.
[0063]
The second embodiment of the present invention shown in FIG. 7 is based on the use of a plurality of information detection means 1a, 1b, 1c,. When one sensor having significant sensitivity with respect to a plurality of degrees of freedom as described with reference to FIG. 2 is used as information detection means, the signal from this sensor includes information corresponding to these degrees of freedom, Since a signal that can be described by one signal waveform can be obtained, information on each body motion mode can be extracted from the obtained signal waveform by correlation evaluation or the like. On the other hand, when a plurality of information detection means are used, it is necessary to perform correlation evaluation on each signal, so that the burden of signal processing increases. However, on the other hand, there is a demand for actually using a combination of a plurality of existing sensors.
[0064]
The second embodiment of the present invention can satisfy such a requirement. As shown in FIG. 7, signals obtained from a plurality of information detecting means 1a, 1b, 1c,. The signal synthesizer 11 superimposes the signal on one signal in advance. Since the number of signals to be processed becomes one by the superposition of signals, the scale of signal processing does not increase according to the number of information detection means. Also, by using correlation evaluation for signal processing, necessary information can be extracted later from the superimposed signal, so that necessary information before superposition can be prevented from being lost. In addition, as is apparent from FIG. 7, the second embodiment of the present invention obtains the detection signal 101 by signal superposition, and then performs the same signal processing as in the case of the first embodiment of the present invention described above. Therefore, the same system configuration as that of the first embodiment can be adopted regardless of the configuration of the information detection means.
[0065]
The signal superimposing means 11 is simply configured to add the signals obtained from the individual information detecting means 1a, 1b, 1c,... And normalize and output the maximum amplitude value after the addition. In addition, different weights may be used for each information detection means at the time of addition. Further, instead of adding the signals as they are, addition by decibel values, that is, logarithmic values of the respective signals may be added. Furthermore, the signals can be added after arbitrary function transformation. More generally, the output 101 may be determined in the form of a multivariable function using the value of each signal as an argument. Specifically, not only the output is determined by the arithmetic processing, but also an arbitrary value may be sequentially output by referring to the map. The signal superimposing method by the signal superimposing means 11 can employ various synthesizing methods depending on the scale allowed for signal processing, but it is important to perform synthesizing without losing the waveform characteristics of each signal. By devising the way of synthesis, there are cases where synthesis that emphasizes signal changes can be performed.
[0066]
The signal processing means 3 may extract a part of the signals without superimposing and use them for signal processing, if necessary, if the signal processing load is sufficient. One example will be described. Assume that the information detection units 1a, 1b, and 1c are acceleration sensors having sensitivity in the x, y, and z axis directions, respectively. The signal synthesizer 11 calculates the sum of the squares of the outputs, takes the square root thereof, and determines the magnitude of the acceleration vector. In addition to this, the signal synthesis means 11 can determine the direction of the acceleration vector by calculating the direction cosine of the vector by another signal processing not shown in the figure. The signal processing means 3 uses only the magnitude of the vector that is the output of the signal synthesizing means 11 for detecting the body motion with relatively coarse accuracy, and for detecting the body motion with finer accuracy, Information regarding the direction of the vector can be referred to as appropriate. By adopting such a method, the second embodiment of the present invention can perform fine signal processing only on necessary portions while reducing the burden of most signal processing, and efficient signal processing. It becomes possible to do. Note that, as the above-described reference information, not only information related to the vector direction but also various information such as rotational angular velocity can be combined.
[0067]
FIG. 8 is a block diagram showing a configuration of a body motion sensing system according to the third embodiment of the present invention. In FIG. 8, 36 is a mounting position self-determination processing unit, 37 is a reference body motion mode detection processing unit, 38 is a selection / collation processing unit, 39 is a determination / output processing unit, and other symbols are those in FIG. Is the same.
[0068]
In the first and second embodiments of the present invention described above, the mounting position information input to the information input unit 21 of the information input unit 2 has been described as being manually input. In general, it is difficult for a user to select a portable body motion sensing system because it is difficult to determine whether the body motion sensing system is mounted on a body or in a bag. However, for example, a device that is initially selected to be placed in a breast pocket may be rolled into the bag and used continuously. In such a case, if the position information can be automatically set, the usability can be improved.
[0069]
In the third embodiment of the present invention, position information can be automatically set. As shown in FIG. 8, the mounting position self-determination unit 36 is provided inside the signal processing means 3. Is. The mounting position self-determination processing unit 36 includes a basic body movement mode detection processing unit 37 that detects a reference body movement mode, a selection / collation processing unit 38 that sequentially selects and collates information stored in the database 35 in advance, and collation As a result, the current system mounting position is determined, and if the result does not match the mounting position information stored in the storage unit 23, the determination / output processing unit 39 sends an output signal 110 for changing the contents of the storage unit 23. It consists of. Here, the reference body motion mode is a body motion mode that occurs relatively frequently in the activity of the target person and has good reproducibility, for example, a walking state. The database 35 stores reference body motion mode data when the body motion sensing system is carried to a representative position. When the reference body motion mode detection processing unit 37 detects the reference body motion mode during the operation of the system, the selection / collation processing unit 38 refers to the above-mentioned reference body motion mode data and the data corresponding to each mounting position. Assess the magnitude of the correlation. As a result, the determination / output processing unit 39 estimates the current carrying position with the mounting position having a significant intensity with respect to the noise level and having the greatest correlation. Naturally, if correct information is always input to the position information input unit 2, the result of the determination / output processing unit 39 matches the content stored in the storage unit 23. However, if they are different from each other, the determination / output processing unit 39 determines that the input information is different, and changes the content of the storage unit 23 to the content estimated to be correct.
[0070]
The mounting position self-determination process as described above may be performed every time the reference detection mode is detected, or may be performed every certain number of times of detection. Alternatively, the mounting position self-determination process may be automatically executed when a large vibration corresponding to the replacement of the apparatus is detected. Further, the change 110 of the stored content by the determination / output processing unit 39 may be performed every time the above-described inconsistency is detected, or may be performed every certain number of times of detection. If the most likely mounting position information is input to the storage means 23 before the operation of the apparatus, by using the method described above, manual input from the position information input means 2 is completely eliminated. It is also possible to operate the system while self-correcting information. In this case, it is also possible to input correct position information to the storage means 23 for each operation by combining learning functions. As the standard body motion mode detection processing unit 37, a normal body motion mode detection processing unit itself can be used. Further, when the normal body movement mode is detected and it is determined that the body movement mode is the reference body movement mode, the above-described self-determination of the mounting position may be performed every predetermined number of times of detection.
[0071]
FIG. 9 is a block diagram showing a configuration of a body motion sensing system according to the fourth embodiment of the present invention. In FIG. 9, 8 is an auxiliary information detecting device, 81 is a biological information detecting means, 82 is a biological information processing means, 83 is a display / communication means, 84 is a power supply unit, and other reference numerals are the same as those in FIG. It is.
[0072]
In each of the embodiments of the present invention described above, the information detection means 1, the signal processing means 3, and the position information input means 2 for inputting the mounting position information of the information detection means 1 are integrated at substantially the same position. It was configured as a mobile device 7 mounted. However, for example, when the portable device 7 is to be applied to a system related to health care, there is a demand for detecting not only body movement but also various biological information correlated with body movement. For example, it is possible to know in detail the physiological state of the body by simultaneously measuring the pulse, blood pressure, electrocardiogram, body temperature, gas components related to respiration, and the like. These detections are not performed anywhere in the human body. For example, if a pulse is detected, it is necessary to provide a detection unit at a position where the pulse can be taken. There is. The fourth embodiment of the present invention shown in FIG. 9 is an example in which another auxiliary information detection device 8 for this purpose is provided in addition to the portable device 7 so that biological information can also be detected. is there. In the example shown in FIG. 9, the auxiliary information detection device 8 is configured as a watch-like form that can be mounted on an arm. The shape of the auxiliary information detection device 8 is not limited to this.
[0073]
The auxiliary information detection device 8 includes a biological information detection unit 81, a biological information signal processing unit 82 that performs processing on the detection signal 111 detected by the biological information detection unit 81, and a post-processing from the biological information signal processing unit 82. The display / communication means 83 that performs display / communication related to the input / output of the output signal 112 and the reference signal 113 for signal processing, and the power supply unit 84 are configured. Similarly to the case described with reference to FIG. 1, the power supply unit 84 can receive the power supplied from the outside via the display / communication means 83 as the power supply power 202. The display / communication means 83 may have a display means for displaying the output result as the name suggests. The apparatus 8 shown in FIG. 9 has a display unit as schematically shown.
[0074]
The biological information detection means 81 is a pressure detection means if it detects a pulse or blood pressure, an electrode means for detection if an electrocardiogram is taken, and a thermopile, thermocouple, infrared sensor if it detects body temperature. The temperature detecting means such as can be configured using a gas detecting means such as a carbon dioxide sensor if the component analysis of the gas related to breathing is performed. In FIG. 9, the biometric information detection means is collectively shown as biometric information detection means 81. In addition, the biological information detection means 81 measures the pulse at the wrist using a wristwatch-like device, and at the same time detects gas related to breathing using a necklace-like device. A plurality of detection means may be provided. When a plurality of pieces of biological information are detected, which biological information is to be detected may be selected by a reference signal 113 from the outside. In this case, the information requested by communication from the portable device 7 or from the outside of the portable device 7 via the portable device 7 can be output again by communication for each device.
[0075]
The signal processing means 82 extracts further necessary information from the output signal of the lower biological information detection means 81. The required information is the change in pulse rate and blood pressure per unit time if a pulse or blood pressure is detected, or a digital value obtained by A / D conversion. If an electrocardiogram is taken, the potential change or A / D converted digital value. If body temperature is detected, it is a temperature value calculated based on thermopile, thermocouple electromotive force and infrared sensor output. For example, the concentration of carbon dioxide, the digital value obtained by A / D conversion of the carbon dioxide, or the number of breaths obtained from the periodic change in the concentration.
[0076]
The display / communication unit 83 transmits the information processed by the signal processing unit 82 to the portable device 7 in a wired or wireless form. Conversely, the display / communication means 83 can obtain body movement information of the portable device 7 and information obtained from the outside by the portable device 7 via the display / communication means 4 of the portable device 7. Since the body motion information and the biological information are correlated with each other, the signal processing performed by the signal processing unit 82 of the auxiliary information detection device 8 and the signal processing performed by the signal processing unit 3 of the portable device 7 are linked to each other to generate a signal. It can also be processed. For example, even with the same bending and stretching exercises, if the number of pulses and breathing increases, it can be seen that the exercise is heavy, such as lifting a heavy load on the shelf. Further, third information that helps signal processing can be acquired from the outside and used for signal processing. The third information is, for example, information such as where the person to be measured is located, and if there is a person in the library room who is doing a heavy exercise such as lifting luggage on the shelf as described above, It turns out that bookshelves are being organized, and in particular, it can be estimated that bookshelves are being organized for heavy books such as dictionaries and books that have already been bound. I understand that. Furthermore, if it is simply a repetitive flexion and extension movement, it can be estimated that the bookcase being organized is not high enough to be lifted using a ladder.
[0077]
The auxiliary information detection device 8 that is an auxiliary information detection means is not limited to the biological information as described above, and may detect a body movement in the same manner as the portable device 7. For example, the movement of the arm can be detected by attaching the device 8 to the arm, and the details of the body movement can be analyzed in the portable device 7 while referring to the result. The body motion sensing system according to the embodiment of the present invention includes the mounting position input unit 2 and the processing unit that performs signal processing based on the mounting position input unit 2 in the portable device 7, and thus is more accurate according to various portable positions of the portable device 7. Although body motion information can be obtained, more detailed or more reliable body motion information can be extracted by combining information obtained from the auxiliary information detection device 8 at a fixed position.
[0078]
The communication of information between the device 7 and the device 8 can be implemented in a wired or wireless form, but in particular, by using the wireless form as shown symbolically by the arrow in FIG. It is preferable without restricting movement. In this case, the wireless reachable range only needs to be able to communicate with each other in the vicinity, for example, between the human arm and the waist, and it is not necessary to consume a large amount of power for communication. Various information can be targeted as communicable information. In particular, frequency information that can be described by the number of events that occur within a certain time interval, such as respiratory rate, heart rate, body movement rate, etc., has a noise-resistant nature as the form of information, so it is subject to measurement and communication. Suitable as For example, in the case of continuously measuring changes in body temperature, if the detected voltage signal is disturbed by noise during detection or communication, information on the resulting body temperature also changes. On the other hand, if the voltage signal is disturbed by noise, correct frequency information can be obtained as long as the number of peaks of the unit time value does not change. Can communicate. On the other hand, analog information such as the body temperature information is preferably converted from voltage signals to temperatures in advance and communicated as digital data.
[0079]
FIG. 10 is a diagram for explaining element integration of the portable device 7, which will be described below. In FIG. 10, 41 is a display means, 42 is a communication means, 68 is an integrated element, and other numerals are the same as those in FIG.
[0080]
The portable device 7 or the auxiliary information detection device 8 described in the embodiments of the present invention described above is preferably a small device having portability, and is compact by integrating internal functions into one chip. Implementation can be realized. The example illustrated in FIG. 10 describes an internal configuration in the case where the basic functions of the mobile device 7 are integrated. The same applies to the device 8.
[0081]
In the example shown in FIG. 10, the information detection means 1, signal processing means 3, communication means (I / O) 42, and position information control unit 22 are integrated on a silicon substrate as shown by hatching. Assume that you want to. As described with reference to FIG. 2, a sensor provided by applying a semiconductor process technology on a silicon substrate can be used as the information detection unit 1. The structure of a sensor such as a diaphragm can be formed by stacking polycrystalline silicon, and the electrodes and wiring portions can be formed by providing a pattern made of aluminum or the like. The electrical insulation can be configured using a silicon oxide film or a nitride film. In order to obtain a hollow structure, a layer called a sacrificial layer such as an oxide is once deposited, then polycrystalline silicon is stacked, and then the sacrificial layer is removed by etching. By using such surface micromachine technology, in addition to a pressure sensor, an acceleration sensor, and an angular velocity sensor, a gas sensor or the like can be made on-chip by a combination with a temperature sensor or a functional material. In addition, the signal processing means 3, the communication means (I / O) 42, and the position information control section 22 are analog including an A / D conversion means, a storage means (ROM, RAM), and a protection circuit, which are also configured using a semiconductor process. A circuit, a digital circuit, and D / A conversion means can be combined. In this case, necessary circuits and memories can be configured based on the CMOS or BiCMOS process technology.
[0082]
As described above, all the functions of the sensor and the circuit can be configured on, for example, a silicon chip having a size of 3 mm-5 mm square and a thickness of about 0.5 mm. In order to exchange signals with the outside, an electrode pad provided around the chip and an external electrode terminal may be electrically connected by a technique such as wire bonding. The integration method is not limited to the above-described method of combining a silicon substrate and a semiconductor process, and a method of integrating and mounting each functional element on a small circuit board can also be adopted. In this case, the position information input unit 21, the display unit 41, the power supply unit 5 and the like can be integrated and mounted on one substrate. As described above, by using the integrated element 68 and the integrated substrate, the apparatus can be easily reduced in size, and the mass productivity can be improved by integrating the main components into one chip.
[0083]
FIG. 11 is a diagram for explaining the configuration of a power generating element that can be used as a power supply unit. This will be described below. In FIG. 11, 50 is a silicon chip, 51 is a mass body, 52 is a support means (beam), 53a, 53b, 53c, and 53d are conductive portions, 54 is an electrode pad, 55 is a glass substrate, and 56 is a permanent magnet.
[0084]
A chargeable / dischargeable battery can be used for the power supply unit 5 of the portable device 7 and the power supply unit 84 of the device 8. In addition, as described above, the power may be supplied from the outside by wire or wireless. On the other hand, some applications of body motion sensing systems must be used in a maintenance-free state for a long time. For example, when trying to observe the ecological behavior of animals, it is desirable to use a device that is as small as possible and can be used for a long time. In such a case, a power generation function may be provided in the body motion sensing system. As an energy supply source that is the basis of power generation, there is a target that uses external energy such as solar cells, temperature difference between body temperature and outside air, or uses vibration energy of body movement itself. There are some that recover and use the energy generated. The example shown in FIG. 11 is of a type that collects vibration energy of body movement, and is particularly a small power generation system that can serve as a power source when an extremely small and low power consumption body movement sensing system is made in the future. In order to make the explanation easy to understand, FIG. 11 shows a state where it is mounted on a package alone. However, as described below, it can be basically configured by processing a silicon substrate. It can also be implemented as part of the integrated device described in FIG.
[0085]
The example shown in FIG. 11 shows a mass body (movable part) 51 obtained by processing a silicon chip on a silicon chip 50 cut out from a silicon substrate, and a support means for supporting the movable part on a part of the silicon chip as well. (Beam) 52 is provided, and a permanent magnet 56 is provided below the mass body (movable part) 51. The movable portion 51 is provided with conductive portions 53a, 53b, 53c, and 53d. In the above description, when vibration associated with body movement is applied to the mass body 51 supported by the support means (beam) 52, the mass body 51 starts to vibrate within the silicon chip surface. This vibration is determined by the mechanical vibration characteristics of the support means 52 as a beam. For example, if the width of the beam is narrowed and the length of the beam is increased, the rigidity as the beam is lowered and the spring becomes soft, so that the resonance frequency is lowered. On the contrary, if the width of the beam is increased and the length of the beam is shortened, the rigidity as the beam is increased and a hard spring is formed, so that the resonance frequency is increased. That is, the mechanical vibration characteristics can be controlled by the design of the support means 52.
[0086]
Here, in the mode of body motion to be measured by this body motion sensing system, in the mode with the highest occurrence frequency, the peak of the vibration frequency of the body motion applied to the power generation means, If the resonance frequency is matched, vibration of the mass body 51 can be excited in a normal use environment of the body motion sensing system. The mass body 51 vibrates in the magnetic field generated by the permanent magnet 55 as the magnetic field generating means. The conductive portions 53a, 53b, 53c, 53d such as aluminum patterned on the surface of the mass body 51 vibrate so as to cross the magnetic field. As a result, the conductive portions 53a, 53b, 53c, 53d are accompanied by vibration. AC voltage is induced. In FIG. 11, these are indicated by V1 and V2. V1 represents an electromotive force induced by the vertical vibration of the mass body 51 in the drawing, and V2 represents an electromotive force induced by the horizontal vibration of the mass body 51 in the drawing. Since the electromotive forces V1 and V2 can be rectified and then stored in a capacitor, a chargeable / dischargeable battery, etc., they can be used as a system power supply or an auxiliary power supply. The electrical signal (power) can be taken out simply by establishing electrical connection with the outside via the electrode pad 54.
[0087]
In the silicon chip 50 described above, an oxide film or nitride film, which is an insulating layer, is formed on a silicon wafer, which is a silicon substrate, and electrodes, conductive parts, and wiring patterns made of aluminum or the like are formed thereon, and a nitride film is formed thereon. After the protective film is formed, the silicon substrate can be manufactured by penetrating through the pattern of the protective film as a mask, leaving the beam and the mass as silicon support means. The penetration processing can be performed using a technique of anisotropic silicon processing such as ICP-RIE (Inductively Coupled Plasma-Reactive Ion Etching). Since the processed silicon substrate is easily broken as it is, for example, a glass substrate 55 such as a glass wafer may be bonded to a surface opposite to the processed surface. For this bonding, for example, a glass-silicon bonding technique known as anodic bonding can be used. At this time, in order not to hinder the movement of the silicon movable portion 52, it is preferable to perform processing such as making a hole in a corresponding portion of the glass substrate. The silicon chip 50 is formed as an individual chip by cutting out a glass substrate by dicing from an anodically bonded wafer. The cut chip can be fixed in a magnetic field by, for example, a method of adhering a permanent magnet 56 to the back surface of the glass substrate 55. Various permanent magnets 56 can be used. For example, when a neodymium magnet or a samarium magnet is used, a strong magnetic field can be obtained with a small magnet.
[0088]
The power generation element described above uses only two-degree-of-freedom vibration as an excitation source, but employs a structure configured in the silicon substrate surface, so that the integrated element described with reference to FIG. Therefore, a functional element having a power generation function can be provided.
[0089]
FIG. 12 is a block diagram showing a configuration of a health management system or work management system which is an application example of the present invention. 12, 7a, 7b, 7c,... Are portable devices, 9 is a base station, 91 is a communication means, 92 is an index conversion processing unit, 93 is a database, 94 is an analysis / evaluation processing unit, and 95 is a display / display unit. A communication means 96 is a power supply unit.
[0090]
In the application example shown in FIG. 12, information transmitted in a wireless form from the portable device 7 as a portable body motion sensing system according to the embodiments of the present invention described above is received by the base station 9 and signal processing is performed. As a result, health management or work management can be implemented. Of course, all health management or work management can be carried out in the body motion sensing system 7. This system is suitable for a personal health management system and the like.
[0091]
The application example shown in FIG. 12 is a configuration in which a single base station 9 manages portable devices 7a, 7b, 7c,. Can be managed. FIG. 12 shows only the basic configuration of the system. For example, the information management portion for storing data for each ID code corresponding to each person and managing the files is understood. For ease of illustration, illustration is omitted.
[0092]
The base station 9 includes a communication unit 91 for the base station 9 to input / output information to / from the outside, and an index conversion processing unit that converts body movement information acquired by communication into information suitable for purposes such as health management and work management. 92, a database 93 necessary for index conversion, an analysis / evaluation processing unit 94 that performs analysis / evaluation based on the converted index, and a result thereof, or for communication output to another external device Display / communication means 95 and a power supply unit 96 for operating the base station 9. The communication unit 91 passes information 114 such as body movement taken from the portable devices 7 a, 7 b, 7 c,..., Which is an external device and attached to the body of the managed person, to the index conversion processing unit 92. The index conversion processing unit 92 captures the reference signal 115 with reference to the database 93 for index conversion, and passes the output information 116 after the index conversion to the analysis / evaluation processing unit 94 and the display / communication means 95. The analysis / evaluation processing unit 94 passes the output information 117 after the analysis / evaluation to the display / communication unit 95. The power supply unit 96 supplies power to each device in the base station 9, and supplies power (power) to each of the plurality of external mobile devices 7a, 7b, 7c,. When supplying, the instruction signal 203 is sent to the communication means 91.
[0093]
Information sent from each of the plurality of mobile devices 7a, 7b, 7c,... External to the base station 9 is indistinguishable as to what body movement mode the target is currently in This is body movement information. The communication information from each of the portable devices 7a, 7b, 7c,... To the base station 9 is, for example, body movement information stored at that time or for a certain time in response to a request from the base station 9. is there. Further, the communication ends in response to the request. Further, the body movement information may be transmitted from the portable devices 7a, 7b, 7c,... Together with the ID every predetermined time, and other control codes may be transmitted simultaneously. .
[0094]
The body movement information 114 acquired by the communication means 91 is, for example, coded content that the body movement mode is “1”, “2”, or cannot be distinguished. is there. The index conversion processing unit 92 associates specific index values stored in the database 93 with each coded body movement mode. Here, the value of the index is set according to the purpose of the system to which the body motion sensing system is applied. For example, when the purpose of the system is health management, it is possible to know how to move muscles according to each body movement mode. Therefore, if the energy consumption is stored as an index value, the result of conversion by the index conversion processing unit 92 The energy consumption according to the body movement mode is known. The analysis / evaluation processing unit 94 can calculate the calorie consumption by summing the energy consumption using the result of conversion in the index conversion processing unit 92.
[0095]
In the above description, a correlation may be found between the appearance of each body motion mode and the occurrence rate of a specific disease, and the risk of occurrence of disease and the safety level of prevention based on the correlation may be provided in the database 93 as an index. In this case, if the exercise item that contributes to health promotion is set to a positive value, and the case where there is a possibility of damaging health is set to a negative value, the analysis / evaluation processing unit 94 adds the index values to calculate the health level. Can be expressed. In this case, a positive value indicates a healthy state (a state with a low incidence of disease). If you exercise too much in the same exercise, you may be detrimental to your health. In such a case, the analysis / evaluation processing unit 94 not only simply adds the index value but also estimates the total exercise time, and can perform processing such as correcting the output 117 in consideration of the result. .
[0096]
On the other hand, if the purpose of the system is work management, for example, by assigning the work load for each body movement mode as an index in the database 93, the work is carried out with a heavy load even in the same time work. Or whether you are working lightly. The analysis / evaluation processing unit 94 can total the workload of each person by, for example, summing the index values. When work with a large load is concentrated on a specific person, the efficiency of the entire work may be reduced. However, the analysis / evaluation processing unit 94, for example, sums the index values and totals the work load of each person, Such a situation can also be determined. The result of the aggregation is transferred from the display / communication unit 95 to an external management system (not shown), for example.
[0097]
This management system can perform work management such as reviewing appropriate work distribution based on the work load of each person calculated by the analysis / evaluation processing unit 94. Of course, there are also cases where the work efficiency is rather improved by repeating the same work to some extent. In this case, the analysis / evaluation processing unit 94 may not only simply add the index value but also estimate the total exercise time and correct the output 117 in consideration of the result. Further, by setting the index value in the database 93 to “1” and the others to “0” for a specific dangerous work, the analysis / evaluation processing unit 94 is obtained as a result of the processing in the index conversion stage 92. When the index value is “1”, it is possible to prompt the management system to promptly warn the worker that there is a danger. The warning method, for example, not only emits a warning sound, but also instructs the portable device 7a, 7b, 7c,. The message may be sent from a speaker integrated with the apparatus. In this case, the speaker can be regarded as an example of display means in the display / communication means 4 of FIG.
[0098]
In the example shown in FIG. 12 described above, the database 93 has been described as storing and using the relationship between each body motion mode and the index value for each target. However, instead of each body motion mode, as a gesture recognition, A specific index may be assigned when an action, for example, body motion modes “1”, “2”, and “3” occur subsequently.
[0099]
Also, out of the body motion modes that should appear uniformly, especially when the frequency of occurrence of limited body motion modes is large, extract the bias as a statistical feature change to grasp the health state and work state You may be able to. When a burden is applied to only one foot even during the same walk, information related to health management and work safety can be extracted, such as a problem with the body or that a specific work is not performed in a correct manner.
[0100]
The health management system or work management system, which is an application example of the present invention, can perform finer management by using the auxiliary information detection unit as described with reference to FIG. For example, if a change in pulse is detected in parallel in a health management system, it becomes easy to quantify the intensity of exercise. In the work management system, for example, in the case of work at a high altitude where the air is thin, the work state can be managed while monitoring the health state of the worker.
[0101]
In addition, the indistinguishable section shown in the database 93 of FIG. 12 is an area where a specific body motion mode cannot be determined. For example, in the above-described health management system, other than the body motion mode previously selected in the database 93 The average energy consumption in the body movement is used as an index value, or the average work load in the body movement other than the body movement mode previously selected in the database 93 is used as the index value in the work management system. By satisfying, it is possible to extract significant information even in a section in which the body movement mode cannot be specified.
[0102]
When considering an actual system application, the body motion mode to be noticed for each application is generally different. For this reason, as the database 93, for example, only the body motion modes to which attention should be paid are extracted and index values are provided. If so, the scale of the database can be reduced. After a series of operations, the index value corresponding to each body movement mode may be changed in real time, such as changing the index value corresponding to the indistinguishable state.
[0103]
Although the health management system and the work management system as application examples of the present invention have been described with reference to FIG. 12, the application of the body motion sensing system according to the present invention is not limited to these. It can also be applied to systems other than humans, such as animal ecology observation and car driving conditions.
[0104]
Further, although the body motion determination in the embodiment of the body motion sensing system according to the present invention has been described focusing on the correlation evaluation with the reference data, the present invention can add a learning function or a neural network as necessary. It is also possible to employ advanced determination processing such as performing determination processing by a network.
[0105]
According to the body motion sensing system according to each of the embodiments of the present invention described above, input is performed manually or automatically with respect to a conventional system that has focused only on the motion near the center of gravity of the body, or is already stored. By performing signal processing based on the mounting position information of the information detecting means, it is possible to grasp the detailed movement of each part of the human body while capturing the body movement at one place. Here, the fine movement indicates not only a movement in the vicinity of the center of gravity of the body but also a movement including a body movement mode with a small change in the center of gravity, such as waving a hand or bending a neck. As signal processing for that purpose, for example, a database set may be selected based on the mounting position information described above.
[0106]
In addition, the body motion sensing system according to each embodiment of the present invention integrally mounts information detection means, signal processing means, and position information input means for inputting mounting position information of the information detection means at substantially the same position. This makes it possible to improve the function without affecting the portability of the system.
[0107]
In addition, the body motion sensing system according to each embodiment of the present invention performs signal processing, particularly known data after superimposing the output signals of the information detecting means corresponding to a plurality of degrees of freedom in software or hardware. Therefore, even if the degree of freedom of measurement to be targeted increases, an increase in load required for signal processing can be suppressed.
[0108]
In addition, the body motion sensing system according to each embodiment of the present invention supplies power necessary for the system in a wireless manner, or includes a self-power generation unit in a part of the system, thereby Replacement is unnecessary, power supply wiring to the apparatus is unnecessary, and portability of the body motion sensing system can be improved.
[0109]
In addition, the body motion sensing system according to each embodiment of the present invention can integrate information detection means and the signal processing means on a common silicon substrate, and superimpose and synthesize a plurality of pieces of information as one signal for output. By using the information detecting means, it is easy to miniaturize the elements including the signal processing unit, and the apparatus can be mounted in a small size so as to be advantageous for carrying.
[0110]
Further, the body motion sensing system according to each embodiment of the present invention is provided with auxiliary information detection means at a position different from the information detection means, the auxiliary information detection means, and a signal processing unit including the signal processing means, Detecting body movements with freedom in the mobile position of the system by performing wired or wireless information communication between them, and detecting biological information such as pulse, blood pressure, and body temperature that should be measured at specific positions Can be made compatible.
[0111]
Furthermore, since the health management system and work management system using the body motion sensing system according to each embodiment of the present invention can capture body movements that are difficult to reflect in the center of gravity, energy consumption and work associated with body motions The condition and the like can be grasped in detail, and can be reflected in the management of the health condition and the work condition.
[0112]
【The invention's effect】
  As described above, according to the present invention,Even when the mounting position of the body motion sensing device changes, it is possible to provide a body motion sensing device and a body motion sensing system that can grasp the motion of the human body.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a body motion sensing system according to a first embodiment of the present invention.
FIG. 2 is a diagram illustrating a configuration example of an information detection unit 1 provided in the first embodiment of the present invention.
FIG. 3 is a diagram illustrating an example of signal processing according to the first embodiment of the present invention.
FIG. 4 is a diagram illustrating an example of correlation evaluation in a frequency domain.
FIG. 5 is a diagram illustrating an example of correlation evaluation in a time domain.
FIG. 6 is a diagram for explaining an example of a signal waveform characteristic of a body movement mode.
FIG. 7 is a block diagram showing a configuration of a body motion sensing system according to a second embodiment of the present invention.
FIG. 8 is a block diagram showing a configuration of a body motion sensing system according to a third embodiment of the present invention.
FIG. 9 is a block diagram showing a configuration of a body motion sensing system according to a fourth embodiment of the present invention.
FIG. 10 is a diagram illustrating element integration of a portable device.
FIG. 11 is a diagram illustrating a configuration of a power generation element that can be used as a power supply unit.
FIG. 12 is a block diagram showing a configuration of a health management system or work management system that is an application example of the present invention.
[Explanation of symbols]
1, 1a, 1b, 1c Information detecting means
2 Location information input means
3 Signal processing means
4,83 Display / communication means
5, 84 Power supply
7, 7a, 7b, 7c Portable device
8 Auxiliary information detector
9 Base station
11 Signal synthesis means
21 Location information input section
22 Location information input section
23 Memory means
31 A / D conversion processor
32 feature extraction processing unit
32a Correlation evaluation unit
32b Bias detector
33 Body movement determination processing unit
34 Database selection processor
35 Database set
36 Mounting position self-determination processing section
37 Reference body motion mode detection processing unit
38 Selection / Verification Processing Unit
39 Judgment / Output Processing Unit
41 Display means
42 Communication means
50 silicon chips
51 Mass
52 Support means (beam)
53a, 53b, 53c, 53d Conductive part
54 Electrode Pad
55 Glass substrate
56 Permanent magnet
60 cases
61a Pressure detection means
61b Diaphragm
62 Silicon substrate
63 Signal conversion means
64 mass
65 variants
66 Connection means
67 Connector
68 Integrated elements
81 Biological information detection means
82 Biological information processing means
91 Communication means
92 Index conversion processing unit
93 Database
94 Analysis / Evaluation Processing Department
95 Display / communication means
96 Power supply

Claims (12)

In a portable body motion sensing device having at least one information detection means and a signal processing means for obtaining body motion information by processing an output signal of the information detection means,
Mounting position input means for inputting the mounting position of the information detection means;
Said signal processing means, body motion sensing apparatus, characterized by chromatic recognizing body movement recognizing means body movement based on the mounting position information of the information detection means from said mounting position input means. In claim 1,
The body motion sensing device, wherein the information detection means has sensitivity in three axial directions. In claim 1,
The body motion sensing device , wherein the signal processing means includes means for recognizing body motion and a state of a living body based on mounting position information of the information detection means . In claim 1,
A body motion sensing device comprising mounting position estimating means for estimating the mounting position . In claim 1,
The information detection means includes a pressure detection means, a deformable body that covers the pressure detection means, and a mass body that is provided in contact with the deformable body,
The body motion sensing device characterized in that an axis passing through the center or center of gravity of the mass body and orthogonal to the pressure detection surface of the pressure detection means does not pass through the center or center of gravity of the pressure detection surface . In a portable body motion sensing device having at least one information detection means and a signal processing means for obtaining body motion information by processing an output signal of the information detection means,
Mounting position input means for inputting the mounting position of the information detection means;
Body movement recognition means for recognizing body movement based on mounting position information of the information detection means from the mounting position input means;
A database corresponding to the mounting position,
The body motion sensing device recognizes body motion based on a correlation between the database and information from the information detection unit . In claim 6,
A body motion sensing device comprising mounting position estimating means for estimating the mounting position based on correlation between the database and information from the information detecting means . In claim 6,
The information detection means includes a pressure detection means, a deformable body that covers the pressure detection means, and a mass body that is provided in contact with the deformable body,
The body motion sensing device characterized in that an axis passing through the center or center of gravity of the mass body and orthogonal to the pressure detection surface of the pressure detection means does not pass through the center or center of gravity of the pressure detection surface . In claim 6,

Means for calculating a feature amount of detection information of the information detection means;
Means for calculating a correlation between the feature quantity and the database;
A body motion sensing device that recognizes body motion when the correlation is greater than a predetermined threshold value . In a body motion sensing system having information detection means and signal processing means for obtaining body motion information by processing an output signal of the information detection means,

Position information input means for giving mounting position information of the information detection means,
The body motion sensing system characterized in that the signal processing means estimates a motion of a human body part in which the information detection means is not mounted by performing signal processing according to position information from the position information input means. . In claim 10,
A body motion sensing system comprising a health management unit that estimates an influence of the motion estimated by the signal processing unit on a health condition . In claim 10,
A body motion sensing system comprising: a work management unit for estimating an influence of the motion estimated by the signal processing unit on work efficiency .

Images