JP6696693B2 - Rotation deviation amount detection device, object detection sensor, rotation deviation amount detection system, rotation deviation amount detection method, and rotation deviation amount detection program - Google Patents

Description

  Embodiments of the present invention relate to a rotation deviation amount detection device, an object detection sensor, a rotation deviation amount detection system, a rotation deviation amount detection method, and a rotation deviation amount detection program.

  At a site such as a building construction, the object detection sensor may be installed on the ceiling in a state where the object detection sensor rotates and deviates from the installation direction. Hereinafter, the amount by which the object detection sensor rotates and deviates with respect to the orientation in which it should be installed is referred to as “rotational deviation amount”. The operator may visually confirm that there is no rotational deviation of the plurality of object detection sensors installed on the ceiling of the room for each object detection sensor. However, conventionally, there have been cases where it is not possible to detect the rotation deviation amounts of a plurality of object detection sensors for each object detection sensor.

Japanese Patent No. 5336100 JP, 2013-4312, A Japanese Patent No. 5872306

"Indoor / outdoor positioning service", [online], [Search on August 1, 2016], <URL: http://sol.panasonic.biz/cloud-service/positioning/> "Integrating multiple surveillance camera images for a bird's eye view! Mitsubishi Electric reveals new technology", [online], [Search on August 1, 2016], <URL: http://www.rbbtoday.com/article/ 2016/03/10 / 140420.html>

  The problem to be solved by the present invention is to provide a rotation deviation amount detection device, an object detection sensor, a rotation deviation amount detection system, and a rotation deviation amount detection device capable of detecting the rotation deviation amounts of a plurality of object detection sensors for each object detection sensor. A method and a rotation deviation amount detection program are provided.

  The rotation deviation amount detection device of the embodiment has a registration unit and a rotation detection unit. A plurality of object detection sensors that detect an object are installed at a plurality of positions. The registration unit registers the correct answer data and the route regarding the route of the moving object. The rotation detection unit, based on the result of comparison between the correct information and the detection information indicating the result of detection of the object that has moved along the registered path by the plurality of object detection sensors, the object detection sensor for the predetermined direction. The amount of rotation deviation is detected.

The figure which shows the example of a structure of the position detection system of embodiment. FIG. 4 is a diagram showing an example of a mask area according to the embodiment. FIG. 4 is a diagram showing an example of a plurality of divided areas defined in a camera image of the embodiment. The figure which shows the example of the map of embodiment. The figure which shows the example of the registration of the route of embodiment. The figure which shows the example of the order in which the object detection sensor etc. detected the pedestrian of embodiment. The figure which shows the example of a detection order of the pedestrian who wears the helmet of embodiment. The figure which shows the 1st example of the position detection result information of the object detection sensor of embodiment. The figure which shows the 2nd example of the position detection result information of the object detection sensor of embodiment. FIG. 6 is a diagram showing an example of a path of movement of a pedestrian in a camera image after distortion correction with a small amount of rotational deviation according to the embodiment. The figure which shows the 1st example of detection of the amount of rotation gaps of embodiment. FIG. 5 is a diagram showing an example of a path of movement of a pedestrian in a camera image after distortion correction in which the amount of rotational deviation according to the embodiment is equal to or greater than a threshold. The figure which shows the 2nd example of detection of the amount of rotation gaps of embodiment. The figure which shows the 3rd example of detection of the rotation shift amount of embodiment. The figure which shows the 4th example of detection of the amount of rotation gaps of embodiment. FIG. 6 is a diagram showing an example of a map display and a list of camera images before distortion correction according to the embodiment. FIG. 4 is a diagram showing an example of map display and cyclic display of camera images before distortion correction according to the embodiment. FIG. 6 is a diagram showing an example of display of a map and selective display of a camera image before distortion correction according to the embodiment. 6A and 6B are diagrams showing an example of display of a map of the embodiment and display of a camera image before distortion correction superimposed on the map. The figure which shows the 1st example of the display of the map of embodiment, and the list display of the camera image after distortion correction superimposed on the map. The figure which shows the 2nd example of the display of the map of embodiment, and the list display of the camera image after distortion correction superimposed on the map. The figure which shows the 3rd example of the display of the map of embodiment, and the list display of the camera image after distortion correction superimposed on the map. 6 is a flowchart showing an example of an operation of detecting the position of the object detection sensor of the embodiment. 6 is a flowchart showing an example of an operation of detecting a rotation deviation amount of the object detection sensor of the embodiment.

  Hereinafter, a rotation deviation amount detection device, an object detection sensor, a rotation deviation amount detection system, a rotation deviation amount detection method, and a rotation deviation amount detection program of an embodiment will be described with reference to the drawings.

  FIG. 1 is a diagram showing an example of the configuration of the position detection system 1. The position detection system 1 is a system that detects the positions of a plurality of object detection sensors 2 installed on the ceiling in a room. The position detection system 1 may be a system (rotational shift amount detection system) that detects the rotational shift amount of the object detection sensor 2. The position detection system 1 includes object detection sensors 2-1 to 2-N (N is an integer of 2 or more; hereinafter, N is 6 as an example), a position detection device 3, and a communication line 4. The object detection sensor 2 and the position detection device 3 may be integrated.

  The object detection sensor 2 detects an object as a sensor of the sensing system. The object detection sensor 2 is, for example, an image sensor, an infrared sensor such as a thermopile sensor, or an ultrasonic sensor. Hereinafter, the object detection sensor 2 is, for example, an image sensor. The object detection sensors 2-1 to 2-N are installed, for example, on the ceiling in the room. The object detection sensor 2 takes a bird's-eye view from the ceiling to capture an image of the room.

  The position detection device 3 is an information processing device such as a notebook-type or desktop-type personal computer, a tablet terminal, or a smartphone terminal. The position detection device 3 may be a server device such as a building energy management system (BEMS).

  The position detection device 3 detects the positions of the plurality of object detection sensors 2 installed on the ceiling in the room based on the result of the object detection sensor 2 detecting the position of the object (moving body) that has moved on the registered route. (presume. For example, the position detection device 3 detects the positions of the plurality of object detection sensors 2 installed on the ceiling in the room based on the image of the moving object in the camera image captured by the object detection sensor 2 looking down from the ceiling.

  The position detection device 3 detects (estimates) the rotation deviation amount of the plurality of object detection sensors 2 installed on the ceiling in the room based on the result of the object detection sensor 2 detecting the position of the object moving on the registered route. The device (rotational deviation detection device) may be used. For example, the position detection device 3 uses the amount of rotation deviation (rotation) of the object detection sensor 2 installed on the ceiling of the room based on the result of comparison between the path of the moving body in the camera image and the correct answer data (correct answer value) regarding the path. Angle) may be detected. The correct answer data regarding the route is data representing the route (reference route) of the moving body in the camera image output by the object detection sensor 2 when the object detection sensor 2 is installed in the direction in which it should be installed. The data representing the path of the moving body includes, for example, a line representing the path of the moving body, an angle representing the direction of the path of the moving body (reference direction), a position (area) on the camera image at which the moving body is first detected, and a movement. It is represented by a position (area) on the camera image where the body is finally detected. The position detection device 3 may correct the amount of rotation deviation that has occurred in the camera image captured by the object detection sensor 2.

  The communication line 4 is a wired or wireless communication line. The network topology of the communication line 4 is, for example, a star type or a ring type.

Next, the configuration of the object detection sensor 2 will be described.
The object detection sensor 2 includes a camera unit 20, an operation unit 21, a storage unit 22, an image recognition unit 23, an ID setting unit 24, a communication unit 25, and a bus 26.

  Some or all of the image recognition unit 23, the ID setting unit 24, and the communication unit 25 function when a processor such as a CPU (Central Processing Unit) executes a program stored in the storage unit 22, for example. It is realized by software. Further, some or all of these functional units may be realized by hardware such as LSI (Large Scale Integration) and ASIC (Application Specific Integrated Circuit).

  The camera unit 20 captures an image of the room from a position installed on the ceiling via a fisheye lens. The camera unit 20 transmits a camera image obtained by capturing an image of the room to the image recognition unit 23. The camera unit 20 captures an image of a moving body moving in a room from a position installed on the ceiling. The moving body is, for example, a walking worker (hereinafter referred to as “pedestrian”), a working vehicle, or a movable work robot. Hereinafter, the description will be continued assuming that the moving body is a pedestrian.

  The operation unit 21 is an operation device such as a DIP switch (Dual In-line Package switch), an operation key, or a touch panel. The operation unit 21 receives an operation by a worker or the like. The operation by the operator or the like receives, for example, an operation of setting parameter information used for image pickup or the like, or a power supply operation.

  The storage unit 22 is configured using a storage device having a non-volatile storage medium (non-transitory recording medium) such as a magnetic hard disk device or a semiconductor storage device. The semiconductor memory device is, for example, a flash memory. The storage unit 22 stores, for example, parameter information used for image capturing and the like, and identification information (ID) of its own sensor. The storage unit 22 may store detection information such as the brightness of an image of a person in the room, the presence / absence of people in the room, the number of people, the degree of congestion, and the amount of activity. The storage unit 22 may store (record) a plurality of camera images in association with image capturing times.

  The memory | storage part 22 memorize | stores the information (henceforth "specific information") for specifying whether the pedestrian designated as a detection target was detected in the camera image by image recognition, for example. The specific information is predetermined by a worker or the like.

  The specific information is, for example, information on colors and patterns of helmets, work clothes, etc. worn by pedestrians. The information on the color and pattern of the helmet, work clothes, etc. may be information based on the product information of the helmet, work clothes, etc. The communication unit or the like of the position detection device 3 may retrieve product information from the product home page via the Internet based on the designated model number of the helmet, work clothes, or the like. The specific information may be personal authentication information such as a pedestrian's profile and gait. The specific information may be wireless communication information of the mobile terminal held by the pedestrian. The specific information may be a teacher image. For example, the specific information may be image information of a person in a camera image captured by the object detection sensor 2 in advance. The specific information may be gesture information such as a pedestrian shaking a hand. The specific information may be information representing a specific area in the room where the pedestrian moves. The specific information may be information representing a character string or the like written on a marker such as a plate worn by a pedestrian.

  The image recognition unit 23 performs image processing on the camera image. For example, the image recognition unit 23 extracts a feature amount such as a luminance change such as a cumulative difference or a luminance distribution of CoHOG (Co-occurrence Histograms of Oriented Gradients) from a camera image.

  The image recognition unit 23 acquires the specific information from the storage unit 22. The image recognition unit 23 may acquire the specific information from the position detection device 3 via the communication unit 25. The image recognition unit 23 specifies whether or not the specified pedestrian is detected in the camera image based on the result of performing the image processing on the camera image and the specific information. For example, the image recognition unit 23 identifies a specified pedestrian on the basis of the processing result of the feature amount by the neural network, the support vector machine (SVM: Support Vector Machine), the k-nearest neighbor discriminator, the Bayes classification, and the specific information. It is specified whether or not it is detected in the camera image.

  At a construction site such as a building, a plurality of workers may work in the same room. Therefore, the image recognition unit 23 needs to identify the pedestrian to be detected in the camera image in which a plurality of pedestrians are recognized. In the camera image in which a plurality of pedestrians are recognized, the image recognition unit 23 determines the pedestrian to be the target of detection and the other pedestrian not to be the target of detection to the specific information designated by the position detection device 3. Identify based on.

  The image recognition unit 23 generates detection information such as the brightness of an image of a person in the room, the presence / absence of people in the room, the number of people, the degree of congestion, and the amount of activity of a pedestrian to be detected. The image recognition unit 23 determines the route of movement of the pedestrian to be detected based on the detection information. The image recognition unit 23 determines the path of movement of the pedestrian to be detected by, for example, background difference processing, interframe difference processing, optical flow processing, and template matching processing. The image recognition unit 23 may add the pedestrian identification information to the detection information.

  Hereinafter, an operation mode (adjustment mode) in which the position detection device 3 detects the position of the object detection sensor 2 is referred to as a “position detection mode”. Hereinafter, the operation mode (adjustment mode) in which the position detection device 3 detects the rotation deviation amount of the object detection sensor 2 is referred to as a “rotation amount detection mode”. Hereinafter, the operation mode of the position detection device 3 other than the position detection mode and the rotation amount detection mode (the operation mode in which the object detection sensor 2 is operated as an image sensor) is referred to as “operation mode”.

  The image recognition unit 23 acquires parameter information used for image capturing or the like from the operation unit 21 or the storage unit 22. The image recognition unit 23 may acquire parameter information used for image capturing or the like from the position detection device 3 via the communication unit 25.

  FIG. 2 is a diagram showing an example of the mask region 100. The parameter information used for imaging or the like is, for example, mask area information for the camera unit 20 of the object detection sensor 2 to mask a partial area of the camera image. When the position detection device 3 operates in the position detection mode, the image recognition unit 23 or the position detection device 3 increases the area of the mask region 100 and decreases the area of the camera image region 110 based on the parameter information. Only the area directly below the object detection sensor 2 may be the pedestrian detection range.

  FIG. 3 is a diagram showing an example of a plurality of areas 120 defined in a camera image. The parameter information used for image capturing or the like is, for example, information on a plurality of regions into which the camera image is divided. The image recognition unit 23 or the position detection device 3 may make the image processing performed on the regions 120 different for each region 120 based on the parameter information. In FIG. 3, the camera image is divided into areas 120-1 to 120-9 as an example.

  Returning to FIG. 1, the description of the configuration of the object detection sensor 2 will be continued. The ID setting unit 24 records the identification information of its own sensor in the storage unit 22. The identification information of its own sensor is, for example, a MAC address (Media Access Control address) and an IP address (Internet Protocol address). The ID setting unit 24 may record the identification information acquired via the operation unit 21 in the storage unit 22 as the identification information of its own sensor. The ID setting unit 24 may record the identification information automatically set by DHCP (Dynamic Host Configuration Protocol) or the like in the storage unit 22 as the identification information of its own sensor. The ID setting unit 24 records the identification information of the own sensor in the storage unit 22 at the time of shipment of the own sensor or after the own sensor is installed on the ceiling, for example.

  The communication unit 25, as an acquisition unit and a transmission unit, communicates with the position detection device 3 via the communication line 4 based on communication protocols such as Ethernet (registered trademark), building, wireless, and device control. The communication unit 25 may communicate with another object detection sensor 2. The communication data is, for example, identification information of its own sensor, detection information based on a result of detection of a pedestrian by the object detection sensor 2, and a camera image of a still image or a moving image based on a result of detection of a pedestrian by the object detection sensor 2. .. The communication protocol of the communication unit 25 is, for example, a general-purpose protocol, an industrial protocol, or an original protocol. The communication unit 25 may communicate with a remote controller that is a device that controls the object detection sensor 2.

  The bus 26 is a transmission path that transmits data between the units of the object detection sensor 2.

Next, the configuration of the position detection device 3 will be described.
The position detection device 3 includes a communication unit 30, an operation unit 31, a map setting unit 32, a registration unit 33, an instruction unit 34, a position detection unit 35, a correction unit 36, a rotation detection unit 37, and a display. The control unit 38, the storage unit 39, the display unit 40, and the bus 41 are provided. The position detection device 3 may further include the image recognition unit 42.

  Some or all of the communication unit 30, the map setting unit 32, the registration unit 33, the instruction unit 34, the position detection unit 35, the correction unit 36, the rotation detection unit 37, the display control unit 38, and the image recognition unit 42 are, for example, , A processor such as a CPU is realized by software that functions by executing a program stored in the storage unit 39. Further, some or all of these functional units may be realized by hardware such as LSI and ASIC.

  The communication unit 30 communicates with the object detection sensors 2-1 to 2-N via the communication line 4 based on communication protocols such as Ethernet (registered trademark), building, wireless, and device control. The communication data is, for example, identification information of the object detection sensor 2, detection information based on the result of detection of the pedestrian by the object detection sensor 2, and a still image or a moving camera image based on the result of detection of the pedestrian by the object detection sensor 2. Is. The communication protocol of the communication unit 30 is, for example, a general-purpose protocol, an industrial protocol, or an original protocol. The communication unit 30 may communicate with a remote controller.

  The operation unit 31 is an operation device such as a DIP switch, an operation key, a keyboard, a mouse or a touch panel. The operation unit 31 receives an operation performed by a worker or the like. The operation unit 31 receives, for example, an operation of registering route data of a pedestrian's movement, an operation of registering correct answer data, an operation of setting parameter information used for image pickup, and a power supply operation.

  The map setting unit 32 acquires a map (plan view) representing the arrangement of columns in the room, partition information, and the like. The map may further include an image (ceiling map) showing the position of the object detection sensor 2 installed on the ceiling. The map is represented by image data such as photographs and diagrams.

  FIG. 4 is a diagram showing an example of the map 5. The position display image 7-n (n is an integer of 1 or more and N or less) is an image representing the position of the object detection sensor 2-n. The map setting unit 32 superimposes the position display image 7-n on the position of the object detection sensor 2-n in the map 5. The position where the position display image 7 is superimposed on the map 5 is determined by, for example, an operator who is looking at the map 5 displayed on the display unit 40 operating the operation unit 31. When superimposing the position display image 7 on the map 5, the map setting unit 32 may superimpose the position display image 7 on the position based on the installation interval information on the map 5.

  At this point, the position detection device 3 has not yet estimated which of the object detection sensors 2-n is installed at the position indicated by each position display image 7.

  The position of the object detection sensor 2 is represented by coordinates and section information. In FIG. 2, the position of the object detection sensor 2-1 (position on which the position display image 7-1 is superimposed) is P1 (“A” bay, first pillar). The position of the object detection sensor 2-2 (the position where the position display image 7-2 is superimposed) is P2 (“A” bay, second pillar). The position of the object detection sensor 2-6 (the position where the position display image 7-6 is superimposed) is P6 (“B” bay, third pillar). The position of the object detection sensor 2 may be associated with time information such as the time when the pedestrian moving inside the room is detected by the object detection sensor 2.

  Returning to FIG. 1, the description of the configuration of the position detection device 3 will be continued. The registration unit 33 includes the route data (line image data) of the movement of the pedestrian in the map 5 in the map 5 according to the operation via the operation unit 31. The registration unit 33 registers the map 5 including the route data of the movement of the pedestrian in the storage unit 39.

  FIG. 5 is a diagram showing an example of registration of the route 6. The worker draws the route 6 of the pedestrian's movement on the map 5 by operating the operation unit 31 such as a mouse while looking at the map 5 displayed on the display unit 40. Path 6 has a start point and an end point. The path 6 does not have to be drawn with a single stroke. When the number of routes 6 is plural, the worker may select the route 6 to be used for position detection of the object detection sensor 2 by operating the operation unit 31.

  Before the pedestrian moves inside the room, the worker draws the route 6 on the map 5 so that the route on which the pedestrian is scheduled to move in the room is shown on the map 5. The pedestrian moves in the room along the route 6 drawn on the map 5 before the pedestrian moves in the room.

  A pedestrian may hold a mobile terminal that performs wireless communication when moving in a room. The mobile terminal held by the pedestrian may estimate the route along which the pedestrian has moved indoors using a position estimation technique (see Non-Patent Document 1). The mobile terminal may transmit data (position estimation result) representing the route along which the pedestrian has moved indoors to the registration unit 33 via the communication unit 30.

  The registration unit 33 may detect the deviation of the route based on the position estimation result acquired by the communication unit 30 from the mobile terminal with respect to the route 6 drawn on the map 5 before the pedestrian moves in the room. The registration unit 33 may transmit the movement route information to the mobile terminal via the communication unit 30 based on the detected deviation. The mobile terminal held by the pedestrian displays the travel route information. Accordingly, the registration unit 33 can guide the pedestrians to the moving route information indicating the route 6 so that the pedestrian can move the route 6 drawn in advance on the map 5 in the room.

  The worker may draw the route 6 on the map 5 so as to trace the route along which the pedestrian has moved in the room after the pedestrian has moved in the room. The registration unit 33 may draw the route 6 on the map 5 based on the position estimation result acquired by the communication unit 30 from the mobile terminal held by the pedestrian after the pedestrian moves in the room.

  The registration unit 33 determines in what order the object detection sensors 2-1 to 2-6 detect pedestrians, the detection range of each object detection sensor 2, the position of each object detection sensor 2 and the route 6. You may make a prediction based on the distance. In FIG. 5, the registration unit 33 is arranged in the order in which a pedestrian passes through the detection range of each object detection sensor 2 (object detection sensors 2-6, 2-5, 2-4, 2-1, 2-2, 2-. It is predicted that each object detection sensor 2 detects a pedestrian in the order of 3).

  The registration unit 33 predicts the time when the pedestrian has passed the detection range of the object detection sensor 2 (passage time) based on the passage time of the pedestrian determined at the position of the object detection sensor 2 according to the route 6. May be. In FIG. 5, the time when the pedestrian passes the detection range of the object detection sensor 2-6 is equal to the passing time t1, for example. The time when the pedestrian passes the detection range of the object detection sensor 2-3 is equal to the passing time t6, for example.

  The registration unit 33 estimates the position where the pedestrian is first detected in the camera image in the camera image when the rotation deviation amount of the object detection sensor 2 is small, based on the relative orientation of the object detection sensor 2 with respect to the route 6. To do. Hereinafter, the position estimated to be the position where the pedestrian is first detected in the camera image is referred to as the “first detection estimated position”. The registration unit 33 records the information indicating the first detected estimated position of the pedestrian in the storage unit 39 as one of the correct answer data.

  The registration unit 33 estimates the position where the pedestrian is finally detected in the camera image in the camera image when the rotation deviation amount of the object detection sensor 2 is small, based on the relative orientation of the object detection sensor 2 with respect to the route 6. To do. Hereinafter, the position estimated to be the position at which the pedestrian is detected last in the camera image is referred to as the “final detection estimated position”. The registration unit 33 records the information indicating the last detected estimated position of the pedestrian in the storage unit 39 as one of correct answer data.

  The registration unit 33 sets the area 120 in which the pedestrian is first detected in the camera image when the rotation deviation amount of the object detection sensor 2 is small in the camera image in which the plurality of areas 120 are defined, to the object detection sensor 2 for the route 6. Estimate based on the relative orientation of. Hereinafter, the area estimated to be the area in which the pedestrian is first detected in the camera image is referred to as a “first detection estimation area”. The registration unit 33 records, in the storage unit 39, information representing the first detection estimation area of the pedestrian as one of the correct answer data.

  The registration unit 33 sets the area 120 where the pedestrian is finally detected in the camera image when the rotation deviation amount of the object detection sensor 2 is small in the camera image in which the plurality of areas 120 are defined, to the object detection sensor 2 for the route 6. Estimate based on the relative orientation of. Hereinafter, the region estimated to be the region in which the pedestrian is detected last in the camera image is referred to as the “final detection estimation region”. The registration unit 33 records, in the storage unit 39, information indicating the final detection estimation region of the pedestrian as one of the correct answer data.

  Returning to FIG. 1, the description of the configuration of the position detection device 3 will be continued. The instruction unit 34 instructs the object detection sensor 2 via the communication unit 30 to change the parameter used for detecting a pedestrian depending on whether the position detection unit 35 detects the position of the object detection sensor 2 or not. Send. That is, the instruction unit 34 transmits different parameter information in the position detection mode and the operation mode to the object detection sensor 2 via the communication unit 30. The instruction unit 34 transmits at least one of the operation mode information and the parameter information to the object detection sensor 2 via the communication unit 30. The operation mode information and the parameter information may be information instructed via the operation unit 31 or information prepared in advance. The instruction unit 34 may select the operation mode and the parameter from the operation modes and the parameters prepared in advance based on a predetermined condition. The instruction unit 34 instructs the object detection sensor on the selected operation mode and parameters. The instruction unit 34 may transmit operation mode information including information for switching the sensing function of the object detection sensor 2 to the object detection sensor 2. The sensing function is, for example, a function of switching the detection accuracy of whether the detection accuracy of the position of the pedestrian in the camera image is pixel unit or area unit.

  For example, when the position detection device 3 operates in the position detection mode, the instruction unit 34 sends parameter information for increasing the area of the mask region 100 as compared with the case where the position detection device 3 operates in the operation mode, and the communication unit 30. You may transmit to the object detection sensor 2 via. For example, when the position detection device 3 operates in the position detection mode, the instruction unit 34 sends parameter information defining a plurality of areas 120 (areas into which the camera image is divided) to the camera image via the communication unit 30 to the object detection sensor. 2 may be sent.

  The instruction unit 34 gives an instruction to change the parameter used for the detection of a pedestrian through the communication unit 30 depending on whether the rotation detection unit 37 detects the rotation deviation amount of the object detection sensor 2 or not. It may be transmitted to the sensor 2. That is, the instruction unit 34 transmits different parameter information in the rotation amount detection mode and the operation mode to the object detection sensor 2 via the communication unit 30.

  The instruction unit 34 acquires the specific information from the operation unit 31 or the storage unit 39. The instruction unit 34 transmits the specific information to the object detection sensor 2 via the communication unit 30. The instruction unit 34 may transmit the specific information to the position detection unit 35 via the bus 41. The instruction unit 34 may transmit the specific information selected from the plurality of specific information based on the operation via the operation unit 31 to the object detection sensor 2 and the position detection unit 35.

  Returning to FIG. 1, the description of the configuration of the position detection device 3 will be continued. The position detection unit 35 detects the position of the object detection sensor 2 for each object detection sensor 2 by associating (associating) the position information of the object detection sensor 2 with the identification information of the object detection sensor 2.

  FIG. 6 is a diagram showing an example of the order in which the object detection sensor 2 or the like detects the pedestrian 130 (moving body). The image illustrated in FIG. 6 is an image based on the camera images transmitted by the plurality of object detection sensors 2. The image illustrated in FIG. 6 is displayed on the display unit 40.

  The pedestrian 130 moves inside the room along the route 6 when the position detection device 3 operates in the position detection mode. In FIG. 6, the position of the pedestrian changes along the route 6 shown in FIG. 5 in the order of the pedestrians 130-1 to 130-6 in time series.

  The detection order display image 8 is an image representing the detection order of the pedestrian 130 "one of the first to sixth". For example, the detection order display image 8-1 represents the detection order “first” of the pedestrian 130 based on that the object detection sensor 2-6 detects the pedestrian 130-1 at the time t1. For example, the detection order display image 8-3 is based on that the object detection sensor 2-3 detects the pedestrian 130-6 at the time t6 after the time t1 and the detection order “6th” of the pedestrian 130. Represents.

  FIG. 7 is a diagram illustrating an example of a detection order of a pedestrian 131 (moving body) wearing a helmet. The image illustrated in FIG. 7 is an image based on the camera images transmitted by the plurality of object detection sensors 2. The image illustrated in FIG. 7 is displayed on the display unit 40.

  The pedestrian 131 moves inside the room along the route 6 shown in FIG. 5 when the position detection device 3 operates in the position detection mode. In FIG. 7, the position of the pedestrian changes in the order of the pedestrians 131-1 to 131-6 in time series along the route 6 shown in FIG. The position detection unit 35 can distinguish the pedestrian 131 from other pedestrians based on the color or pattern of the helmet worn by the pedestrian 131.

  The detection order display image 8 is an image showing the detection order of the pedestrian 131 “any one of Nos. 1 to 6”. For example, the detection order display image 8-1 represents the detection order “first” of the pedestrian 131 based on that the object detection sensor 2-6 detects the pedestrian 131-1 at the time t1. For example, the detection order display image 8-3 is based on that the object detection sensor 2-3 detects the pedestrian 131-6 at the time t6 after the time t1, and the detection order “6th” of the pedestrian 131. Represents.

  The position detection unit 35 (link unit) uses the object detection sensor 2 based on the correspondence relationship between the order in which the object detection sensor 2 detects the pedestrian 131 and the order in which the pedestrian 131 has passed the detection range of the object detection sensor 2. The position of is detected for each object detection sensor 2.

  FIG. 8 is a diagram showing a first example of the position detection result information of the object detection sensor 2. In FIG. 8, the pedestrian detection order, the identification information of the object detection sensor 2, the order in which the pedestrian passed the detection range of the object detection sensor 2 (the pedestrian's passing order), and the position of the object detection sensor 2. Are associated with.

  The position detection unit 35 associates the identification information of the object detection sensor 2 that has detected the pedestrian 131 with the order of detection of the pedestrian 131 and registers it in the position detection result information. For example, the position detection unit 35 associates the identification information “2-6” of the object detection sensor 2 that first detects the pedestrian 131 with the detection order “first” of the pedestrian 131, and detects the position detection result information. Register with. For example, the position detection unit 35 associates the identification information “2-3” of the object detection sensor 2 that has detected the pedestrian 131 sixth with the detection order “6” of the pedestrian 131, and detects the position detection result information. Register with.

  The position detection unit 35 acquires the position information of the object detection sensor 2 from the operation unit 31 or the registration unit 33 in association with the order in which the pedestrian 131 has passed the detection range of the object detection sensor 2 (passing order). The position detection unit 35 associates the identification information of the object detection sensor 2 with the position information of the object detection sensor 2 so that the order of detection of the pedestrian 131 and the order of passage of the pedestrian 131 are the same.

  For example, the position detection unit 35 is associated with the identification information "2-6" of the object detection sensor 2 associated with the pedestrian's passing order "1" and the pedestrian's detection order "1". The position “P6” of the object detection sensor 2-6 is associated. For example, the position detection unit 35 is associated with the identification information “2-3” of the object detection sensor 2 associated with the pedestrian passage order “6” and the pedestrian detection order “6”. The position "P3" of the object detection sensor 2-3 is associated.

  Accordingly, the position detection unit 35 can detect the position of the object detection sensor 2 for each object detection sensor 2. For example, the position detection unit 35 can detect that the object detection sensor 2-6 assigned with the identification information “2-6” is installed at the position “P6” in the room. For example, the position detection unit 35 can detect that the object detection sensor 2-3 to which the identification information “2-3” is assigned is installed at the position “P3” in the room.

  The position detection unit 35 (link unit) uses the object detection sensor 2 based on the correspondence between the time when the object detection sensor 2 detects the pedestrian 131 and the time when the pedestrian 131 passes the detection range of the object detection sensor 2. The position of may be detected for each object detection sensor 2.

  FIG. 9 is a diagram showing a second example of the position detection result information of the object detection sensor 2. In FIG. 9, the pedestrian detection time, the identification information of the object detection sensor 2, the time when the pedestrian passed the detection range of the object detection sensor 2 (passage time of the pedestrian), and the position of the object detection sensor 2. Are associated with. The passage time of the pedestrian is set on the route 6.

  The position detection unit 35 registers the identification information of the object detection sensor 2 that has detected the pedestrian 131 in the position detection result information in association with the detection time of the pedestrian 131. For example, the position detection unit 35 associates the identification information “2-6” of the object detection sensor 2 that has detected the pedestrian 131 at the time t1 with the detection time “t1” of the pedestrian 131, and adds the identification information “2-6” to the position detection result information. register. For example, the position detection unit 35 associates the identification information “2-3” of the object detection sensor 2 that has detected the pedestrian 131 at time t6 with the detection time “t6” of the pedestrian 131, and adds the identification information “2-3” to the position detection result information. register.

  The position detection unit 35 acquires the position information of the object detection sensor 2 from the operation unit 31 or the registration unit 33 in association with the time when the pedestrian 131 has passed the detection range of the object detection sensor 2 (passing time). The position detection unit 35 associates the identification information of the object detection sensor 2 with the position information of the object detection sensor 2 so that the detection time of the pedestrian 131 and the passage time of the pedestrian 131 become equal.

  For example, the position detection unit 35 identifies the identification information “2-6” of the object detection sensor 2 associated with the pedestrian passage time “t1” and the object detection associated with the pedestrian detection time “t1”. The position "P6" of the sensor 2-6 is associated. For example, the position detection unit 35 identifies the object detection sensor 2 associated with the passing time “t6” of the pedestrian “2-3” and the object detection associated with the pedestrian detection time “t6”. The position "P3" of the sensor 2-3 is associated.

  Accordingly, the position detection unit 35 can detect the position of the object detection sensor 2 for each object detection sensor 2. For example, the position detection unit 35 can detect that the object detection sensor 2-6 assigned with the identification information “2-6” is installed at the position “P6” in the room. For example, the position detection unit 35 can detect that the object detection sensor 2-3 to which the identification information “2-3” is assigned is installed at the position “P3” in the room.

  Returning to FIG. 1, the description of the configuration of the position detection device 3 will be continued. The correction unit 36 corrects the distortion of the camera image (hereinafter, referred to as “pre-distortion correction camera image”) by the fisheye lens of the camera unit 20 of the object detection sensor 2. The correction unit 36 transmits the camera image after distortion correction (hereinafter referred to as “distortion corrected camera image”) to the rotation detection unit 37 and the display control unit 38.

  When the object detection sensor 2 is installed on the ceiling, a rotation deviation amount may occur in the direction of the object detection sensor 2 with respect to the reference. The correction unit 36 acquires information indicating the orientation of the object detection sensor 2 or the rotation deviation amount of the camera image from the rotation detection unit 37 for each object detection sensor 2. The correction unit 36 corrects the rotation deviation amount of the distortion-corrected camera image based on the rotation deviation amount. The correction unit 36 transmits the distortion-corrected camera image in which the amount of rotation deviation has been corrected to the display control unit 38.

  The rotation detection unit 37 acquires a camera image. The rotation detection unit 37 detects the orientation of the installed object detection sensor 2 or the rotation deviation amount of the camera image based on the comparison result of the camera image and the correct answer data. The rotation detection unit 37 transmits information indicating the orientation of the object detection sensor 2 or the rotation deviation amount of the camera image to the correction unit 36 and the display control unit 38.

  The rotation detection unit 37 may output information indicating the amount of rotation deviation of the installed object detection sensor 2 by voice. When the pedestrian 131 passes the detection range of the object detection sensor 2 whose rotation deviation amount is equal to or more than a threshold value, the rotation detection unit 37 causes the pedestrian 131 to detect the detection range of the object detection sensor 2 whose rotation deviation amount is equal to or more than the threshold value. Information indicating that the vehicle has passed may be notified to the operator or the like by voice or the like. The rotation detection unit 37 may notify the worker or the like of the distance between the position of the object detection sensor 2 and the position of the pedestrian 131, the rotation deviation amount of which is equal to or more than the threshold value, by voice or the like. The rotation detection unit 37 may notify the worker or the like by voice or the like of the direction of the position of the object detection sensor 2 whose rotation deviation amount is equal to or greater than a threshold, with the position of the pedestrian 131 as the origin.

  The rotation detection unit 37 detects the orientation of the installed object detection sensor 2 or the rotation deviation amount of the camera image based on the correspondence relationship between the first area and the second area on the camera image and the movement path of the pedestrian. To do. For example, the rotation detection unit 37 detects the orientation of the installed object detection sensor 2 or the rotation deviation amount of the camera image based on the correspondence relationship between a plurality of areas into which the camera image is divided and the path of movement of the pedestrian. ..

  FIG. 10 is a diagram illustrating an example of a movement path of the pedestrian 131 in the distortion-corrected camera image 140-2 with a small amount of rotation deviation. In the distortion-corrected camera image 140-2, the positions of the pedestrians are the pedestrians 131-5-1, 131-5-2, 131-5-3 in chronological order along the route 6 shown in FIG. Change. Note that the pedestrian 131 does not have to go straight on the distortion-corrected camera image 140-2 as long as the pedestrian 131 is moving along the route 6 shown in FIG.

  FIG. 11 is a diagram showing a first example of detecting the amount of rotation deviation. In FIG. 11, the distortion-corrected camera image 140-2 shown in FIG. 10 is divided into regions 120-1 to 120-9. In FIG. 11, the region (entry region) in which the image recognition unit 23 or the image recognition unit 42 first detects the pedestrian 131 in the camera image 140-2 after distortion correction is the region 120-4.

  When the image recognizing unit 23 or the image recognizing unit 42 detects the pedestrian 131 over the adjacent regions 120, the image recognizing unit 23 or the image recognizing unit 42 may select the region 120 in which the pedestrian 131 is detected in any of the adjacent regions 120. ..

  The region in which the pedestrian 131 is first detected by the image recognition unit 23 or the image recognition unit 42 in the distortion-corrected camera image 140-2 may be the region 120-5 because it takes time to detect the pedestrian 131. The area (exit area) in which the image recognition unit 23 or the image recognition unit 42 finally detects the pedestrian 131 in the distortion-corrected camera image 140-2 is the area 120-6.

  The rotation detection unit 37 acquires, from the registration unit 33 or the storage unit 39, information indicating the region 120-4 that is the first detection estimation region of the pedestrian 131 in the distortion-corrected camera image 140-2. The rotation detection unit 37 acquires, from the registration unit 33 or the storage unit 39, information indicating the region 120-6 that is the final detection estimation region of the pedestrian 131 in the distortion-corrected camera image 140-2.

  The rotation detection unit 37 determines whether or not the region in which the pedestrian 131 is first detected in the post-correction camera image 140-2 and the region 120-4 which is the first detection estimation region are the same region based on the correct answer data. To judge. In the rotation detection unit 37, the region where the image recognition unit 23 or the image recognition unit 42 detects the pedestrian 131 last in the distortion-corrected camera image 140-2 and the region 120-6 that is the final detection estimation region are the same region. Whether or not there is is determined based on the correct answer data.

  In FIG. 11, in the rotation detection unit 37, the region 120 in which the pedestrian 131 is detected first and the first detection estimated region of the pedestrian 131 are the same region, and the region 120 in which the pedestrian 131 is detected last and the pedestrian 131 are the same. Since it is the same area as the last detection estimation area of, the direction of the installed object detection sensor 2 or the rotation deviation amount of the camera image is determined to be small.

  FIG. 12 is a diagram showing an example of the path of movement of the pedestrian 131 in the camera image 140a after distortion correction in which the amount of rotation deviation is equal to or greater than the threshold. In the distortion-corrected camera image 140a-2, the positions of the pedestrians are the pedestrians 131-5-1, 131-5-2, 131-5-3 in chronological order along the route 6 shown in FIG. Change. Note that the pedestrian 131 does not have to go straight in the distortion-corrected camera image 140a-2 as long as the pedestrian 131 is moving along the route 6 shown in FIG.

  FIG. 13 is a diagram showing a second example of detecting the amount of rotation deviation. In FIG. 13, the distortion-corrected camera image 140a-2 shown in FIG. 10 is divided into regions 120-1 to 120-9. In the distortion-corrected camera image 140a-2, the area (entry area) in which the image recognition unit 23 or the image recognition unit 42 first detects the pedestrian 131 is the area 120-1.

  The area in which the image recognizing unit 23 or the image recognizing unit 42 first detects the pedestrian 131 in the distortion-corrected camera image 140a-2 may be the area 120-5 because detection takes time. An area (exit area) in which the image recognition unit 23 or the image recognition unit 42 finally detects the pedestrian 131 in the distortion-corrected camera image 140a-2 is the area 120-9.

  The rotation detection unit 37 determines whether or not the area in which the pedestrian 131 is first detected in the distortion-corrected camera image 140a-2 and the area 120-4 that is the first detection estimation area are the same area. In the rotation detection unit 37, the region where the image recognition unit 23 or the image recognition unit 42 finally detects the pedestrian 131 in the distortion-corrected camera image 140a-2 and the region 120-6 that is the final detection estimation region are the same region. Determine if there is.

  In FIG. 13, when at least one of the area 120 in which the pedestrian 131 is first detected and the area 120 in which the pedestrian 131 is finally detected is not a predetermined area, the rotation detection unit 37 detects that the object detection sensor 2 installed. It is determined that the rotation deviation amount is equal to or more than the threshold value. In the rotation detection unit 37, since the area 120 in which the pedestrian 131 is first detected and the first detection estimation area of the pedestrian 131 are different areas, the orientation of the installed object detection sensor 2 or the rotation deviation amount of the camera image is It is determined that it is equal to or more than the threshold value. In the rotation detection unit 37, since the region 120 in which the pedestrian 131 is detected last and the last detection estimation region of the pedestrian 131 are different regions, the direction of the installed object detection sensor 2 or the rotation deviation amount of the camera image is You may determine that it is more than a threshold value.

  The rotation detection unit 37 may detect the orientation of the installed object detection sensor 2 or the rotation deviation amount of the camera image based on the angle formed by the coordinate axis defined in the camera image and the path of movement of the pedestrian. .. Note that the pedestrian 131 may move along the path 6 so that the angle formed by the coordinate axis defined in the camera image and the path 6 of the movement of the pedestrian 131 is constant. In this case, the correct answer data of the angle indicating the direction (reference direction) of the path of the moving body is registered in advance.

  FIG. 14 is a diagram showing a third example of detecting the amount of rotation deviation. An XY coordinate axis (two-dimensional coordinate axis) is defined in the distortion-corrected camera image 140-2 shown in FIG. In FIG. 14, the X axis indicates the reference direction of the route 6 (the reference direction of movement of a pedestrian) when the amount of rotation deviation of the object detection sensor 2 is small.

  The position where the image recognition unit 23 or the image recognition unit 42 first detects the pedestrian 131 in the camera image 140-2 after distortion correction is the position of the pedestrian 131-5-1. The position where the image recognizing unit 23 or the image recognizing unit 42 finally detects the pedestrian 131 in the distortion-corrected camera image 140-2 is the position of the pedestrian 131-5-3.

  A broken line 200 is a broken line connecting a position where the pedestrian 131 is first detected to a position where the pedestrian 131 is finally detected in the camera image 140-2 after distortion correction. Therefore, the broken line 200 represents the direction of movement of the pedestrian 131 in the camera image 140-2 after distortion correction.

  In FIG. 14, the rotation detection unit 37 indicates that the direction of the route 6 (reference direction of movement of the pedestrian) and the direction of broken line 200 (direction of movement of the pedestrian) when the amount of rotation deviation of the object detection sensor 2 is small. Since they match, it is determined that the orientation of the installed object detection sensor 2 or the amount of rotation deviation of the camera image is small.

  FIG. 15 is a diagram showing a fourth example of detecting the amount of rotation deviation. An XY coordinate axis (two-dimensional coordinate axis) is defined in the distortion-corrected camera image 140a-2 shown in FIG. In FIG. 15, the X axis indicates the reference direction of the route 6 (the reference direction of movement of a pedestrian) when the amount of rotation deviation of the object detection sensor 2 is small (less than the threshold value).

  The position where the image recognizing unit 23 or the image recognizing unit 42 first detects the pedestrian 131 in the distortion-corrected camera image 140a-2 is the position of the pedestrian 131-5-1. The position where the image recognizing unit 23 or the image recognizing unit 42 finally detects the pedestrian 131 in the distortion-corrected camera image 140a-2 is the position of the pedestrian 131-5-3.

  A broken line 200 is a broken line connecting a position where the pedestrian 131 is first detected to a position where the pedestrian 131 is finally detected in the post-correction camera image 140a-2. Therefore, the broken line 200 represents the direction of movement of the pedestrian 131 in the camera image 140a-2 after distortion correction. In FIG. 15, the broken line 200 is inclined (−45 degrees) with respect to the X axis.

  In the rotation detection unit 37 in FIG. 15, the reference direction (X axis) representing the route 6 in the camera image 140a-2 after distortion correction does not coincide with the direction of the broken line 200 (direction of movement of the pedestrian 131). It is determined that the orientation of the installed object detection sensor 2 or the rotation deviation amount of the camera image is equal to or more than a threshold value. The rotation detection unit 37 transmits information indicating the orientation of the object detection sensor 2 or the rotation deviation amount (−45 degrees) of the camera image to the correction unit 36 and the display control unit 38.

  The display control unit 38 includes the map 5, the route 6, the position display image 7, the detection order display image 8, the presence / absence of the rotation deviation amount, the rotation deviation amount, the camera image before the distortion correction, the camera image after the distortion correction, the detection information, and these. The combination is displayed on the display unit 40.

  FIG. 16 is a diagram showing an example of the display of the map 5 and the list display of the camera images 141 before distortion correction. The display control unit 38 displays the map 5 and the list of the camera images 141 before distortion correction on the display unit 40. The display control unit 38 arranges the pre-distortion camera images 141-1 to 141-6 on the screen of the display unit 40 in association with the position of the object detection sensor 2.

  FIG. 17 is a diagram showing an example of the display of the map 5 and the cyclic display of the camera image 141 before distortion correction. When the operation unit 31 is operated, the display control unit 38 selects the pre-distortion correction camera image 141 from the pre-distortion correction camera images 141-1 to 141-5 at regular time intervals. In FIG. 17, the display control unit 38 selects the pre-distortion correction camera image 141-4 and the pre-distortion correction camera image 141-5. The display control unit 38 cyclically displays the selected pre-distortion correction camera image 141 while switching the pre-distortion correction camera image 141 to be selected.

  FIG. 18 is a diagram showing an example of the display of the map 5 and the selective display of the camera image 141 before distortion correction. When the operation unit 31 is operated, the display control unit 38 selects the pre-distortion correction camera image 141 from the pre-distortion correction camera images 141-1 to 141-5 according to the operation. In FIG. 18, the display control unit 38 distorts the object detection sensor 2-2 associated with the position display image 7-2 by clicking the operation unit 31 and selecting the position display image 7-2. The uncorrected camera image 141-2 is selected. The display control unit 38 enlarges and displays the selected pre-distortion correction camera image 141-2.

  FIG. 19 is a diagram showing an example of display of the map 5 and display of the pre-distortion correction camera image 141 superimposed on the map 5. The display control unit 38 displays the map 5 and the list of the camera images 141 before distortion correction on the display unit 40. The display control unit 38 associates the pre-distortion camera images 141-1 to 141-6 with the position display images 7-1 to 7-6 based on the position of the object detection sensor 2 and arranges them on the screen of the display unit 40. To do.

  FIG. 20 is a diagram showing a first example of the display of the map 5 and the list display of the camera images 140 after distortion correction superimposed on the map 5. The display control unit 38 displays the map 5 and a list of the distortion-corrected camera images 140 on the display unit 40. The display control unit 38 associates the camera images 140-1 to 140-6 after distortion correction with the position display images 7-1 to 7-6 based on the position of the object detection sensor 2 and arranges them on the screen of the display unit 40. To do.

  FIG. 21 is a diagram showing a second example of the display of the map 5 and the list display of the camera images 140 after distortion correction superimposed on the map 5. The display control unit 38 generates a bird's-eye view of the room by arranging the plurality of distortion-corrected camera images 140 so as to be adjacent to each other (see Non-Patent Document 2). The display control unit 38 displays the map 5 and the bird's-eye view on the display unit 40.

  FIG. 22 is a diagram showing a third example of the display of the map 5 and the list display of the distortion-corrected camera images 140 superimposed on the map 5. The display control unit 38 acquires, from the correction unit 36, the distortion-corrected camera image 140-2 in which the rotation deviation amount of the distortion-corrected camera image 140a-2 is corrected. The display control unit 38 generates a bird's-eye view of the room by arranging a plurality of distortion-corrected camera images 140 including the distortion-corrected camera image 140-2 in which the rotation deviation amount has been corrected so as to be adjacent to each other. The display control unit 38 displays the map 5 and the bird's-eye view on the display unit 40.

  The display control unit 38 may execute an operation based on an interactive UI (Interactive User Interface) which is a bidirectional user interface. For example, when the worker clicks on the camera image displayed on the display unit 40 using the operation unit 31, the display control unit 38 transmits a signal to the position detection unit 35 to instruct the position detection process to be redone. To do. When the position detection unit 35 acquires a signal instructing to redo the position detection process, the position detection unit 35 redoes the position detection process. The position detection unit 35 may perform the position detection process again after the pedestrian 130 moves in the room, or may perform the position detection process again while the pedestrian 130 is moving in the room.

  Note that the display control unit 38 may always display the diagrams shown in FIGS. 19 to 22 on the display unit 40, or may display the diagrams on the display unit 40 at the timing when the object detection sensor 2 detects the pedestrian 131. May be. The display control unit 38 may display the latest camera image on the display unit 40 before the detection of the pedestrian 131 is completed. The display control unit 38 may display the best shot of the camera image on the display unit 40 after the detection of the pedestrian is completed. The display control unit 38 may popup-display the camera image on the display unit 40 at the timing when the detection of the pedestrian is started. The display control unit 38 may correct the distortion of the camera image before distortion correction. The display control unit 38 may correct the rotation deviation amount of the distortion-corrected camera image based on the rotation deviation amount.

  Returning to FIG. 1, the description of the configuration of the position detection device 3 will be continued. The storage unit 39 is configured using a storage device having a non-volatile storage medium (non-transitory recording medium) such as a magnetic hard disk device or a semiconductor storage device. The storage unit 39 stores the specific information, the detection information, the position detection program, and the position detection result information.

  The display unit 40 is a display device such as a liquid crystal display. The display unit 40 includes a map 5, a route 6, a position display image 7, a detection order display image 8, presence / absence of a rotation deviation amount, a rotation deviation amount, a camera image before distortion correction, a camera image after distortion correction, detection information, and a combination thereof. Is displayed. The display unit 40 may correct the distortion of the camera image before distortion correction. The display unit 40 may correct the rotation deviation amount of the distortion-corrected camera image based on the rotation deviation amount.

  The bus 41 is a transmission path that transmits data between the respective units of the position detection device 3.

  The image recognition unit 23 performs the image processing such as image recognition performed by the image recognition unit 23 of the object detection sensor 2 instead of the object detection sensor 2.

Next, the operation of the position detection system 1 will be described.
FIG. 23 is a flowchart showing an example of the operation of detecting the position of the object detection sensor 2. The communication unit 30 acquires the identification information of the object detection sensor 2 from the plurality of object detection sensors 2 that have detected a pedestrian (step S101). The position detection unit 35 associates the identification information of the object detection sensor 2 with the position of the object detection sensor 2 based on the result of detecting the pedestrian moving along the route 6 (step S102).

  FIG. 24 is a flowchart showing an example of the operation of detecting the amount of rotation deviation of the object detection sensor 2. The communication unit 30 acquires, from the object detection sensor 2, a camera image based on the result of detecting a pedestrian moving along the route 6 (step S201). The rotation detection unit 37 detects the rotation deviation amount of the object detection sensor 2 based on the position where the pedestrian is first detected in the camera image and the position where the pedestrian is finally detected in the camera image (step S202).

  As described above, the position detection device 3 of the embodiment has the registration unit 33, the communication unit 30, and the position detection unit 35. The plurality of object detection sensors 2 are installed at a plurality of positions. The registration unit 33 registers the route 6 in the storage unit 39. The communication unit 30 acquires the detection information indicating the result of the plurality of object detection sensors 2 detecting the pedestrian who has moved along the registered route 6, and the identification information of the object detection sensor 2. The position detection unit 35 associates the identification information of the object detection sensor 2 with the position of the object detection sensor 2 based on the detection information and the result of the pedestrian passing through the detection ranges of the plurality of object detection sensors 2, The position of the object detection sensor 2 is detected.

  With this configuration, the position detection unit 35 determines the identification information of the object detection sensor 2 and the position of the object detection sensor 2 based on the result of the pedestrian passing the detection ranges of the plurality of object detection sensors 2 and the detection information. Correspond. Thereby, the position detection device 3 of the embodiment can detect the positions of the plurality of object detection sensors 2 for each object detection sensor 2.

  The position detection unit 35 of the embodiment detects the pedestrians by the plurality of object detection sensors 2 so that the order in which the pedestrians pass the detection ranges of the plurality of object detection sensors 2 is equal. The identification information and the position of the object detection sensor 2 may be associated with each other. The position detection unit 35 of the embodiment identifies the object detection sensor 2 so that the time when the plurality of object detection sensors detect the pedestrian and the time when the pedestrian passes the detection range of the plurality of object detection sensors are equal. The information and the position of the object detection sensor 2 may be associated with each other.

  As described above, the position detection device 3 according to the embodiment has the registration unit 33 and the rotation detection unit 37. The registration unit 33 registers the correct answer data regarding the moving pedestrian's route 6 and the route 6 in the storage unit 39. The rotation detection unit 37 detects a pedestrian moving along the registered route 6 based on the comparison result of the detection information indicating the detection result of the plurality of object detection sensors 2 and the correct answer data, and determines the predetermined direction. The amount of rotation deviation of the object detection sensor 2 is detected.

  With this configuration, the rotation detection unit 37 determines in advance based on the comparison result of the detection information indicating the result of the detection of the plurality of object detection sensors 2 of the pedestrian moving along the registered route 6 and the correct answer data. The amount of rotation deviation of the object detection sensor 2 with respect to the determined direction is detected. Thereby, the position detection device 3 of the embodiment can detect the rotation deviation amount of the plurality of object detection sensors 2 for each object detection sensor 2.

  The rotation detection unit 37 of the embodiment compares the region 120 on the camera image where the object detection sensor 2 first detected a pedestrian, the region 120 where the object detection sensor 2 last detected a pedestrian, and the correct answer data. The amount of rotation deviation of the object detection sensor 2 may be detected based on the result. In this case, since the position of the moving body is compared with the coarse accuracy (unit) of the area, the accuracy of the position of the moving body such as a pedestrian in the image may be any accuracy such as pixel, area, or three-dimensional coordinate. Good. The rotation detection unit 37 of the embodiment detects the object detection sensor 2 based on the comparison result between the moving direction of the pedestrian on the camera image and the correct answer data indicating the direction (reference direction) of the route 6 on the camera image. The angle of the rotation deviation amount may be detected.

  According to at least one embodiment described above, an object based on the comparison result of the detection information and the correct answer data indicating the result of the plurality of object detection sensors detecting the object that has moved along the registered route. By providing the rotation detection unit that detects the rotation deviation amount of the detection sensor, the rotation deviation amounts of the plurality of object detection sensors can be detected for each object detection sensor.

  Although some embodiments of the present invention have been described above, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. The embodiments and their modifications are included in the scope of the invention and the scope thereof, and are included in the invention described in the claims and the scope of equivalents thereof.

DESCRIPTION OF SYMBOLS 1 ... Position detection system, 2 ... Object detection sensor, 3 ... Position detection device, 4 ... Communication line, 5 ... Map, 6 ... Route, 7 ... Position display image, 8 ... Detection order display image, 20 ... Camera part, 21 ... operation unit, 22 ... storage unit, 23 ... image recognition unit, 24 ... ID setting unit, 25 ... communication unit, 26 ... bus, 30 ... communication unit, 31 ... operation unit, 32 ... map setting unit, 33 ... registration unit , 34 ... Instructing section, 35 ... Position detecting section, 36 ... Correction section, 37 ... Rotation detecting section, 38 ... Display control section, 39 ... Storage section, 40 ... Display section, 41 ... Bus, 42 ... Image recognition section, 100 ... mask area, 110 ... camera image area, 120 ... area, 130 ... pedestrian, 131 ... pedestrian, 140 ... camera image after distortion correction, 141 ... camera image before distortion correction, 200 ... broken line

Claims (14)

A plurality of object detection sensors for detecting an object are installed at a plurality of positions, a rotation deviation amount detection device for detecting a rotation deviation amount of the object detection sensor,
A registration unit that registers the correct answer data regarding the path of the moving object and the path;
Based on the result of comparison between the correct information and the detection information representing the result of detection of the object moved along the registered path by the plurality of object detection sensors, the object detection sensor of the predetermined orientation A rotation detector that detects the amount of rotation deviation,
Rotational deviation amount detection device comprising: The rotation detecting unit, the object detection sensor is a result of comparison between the position and the correct data on the image position and the object detection sensor detects the object at the end of the images obtained by detecting the object in the first The rotation deviation amount detecting device according to claim 1, wherein the rotation deviation amount of the object detection sensor with respect to a predetermined direction is detected based on the above. The rotation detection part, based on the object detection sensor is a result of comparison between the first region on the images has been detected with the object detection sensor detects the object at the end region with the correct answer data to said object, The rotation deviation amount detection device according to claim 1 or 2, which detects a rotation deviation amount of the object detection sensor. The rotation detecting unit, based on a comparison result of the direction and the correct answer data of the movement of the object on the images, detects the angle of rotation displacement amount of the object detection sensor, to claim 1 or claim 2 The rotation deviation amount detection device described. An image recognition unit is further provided for specifying whether or not the specified object is detected by image recognition,
When the image recognition unit identifies that the rotation detection unit has detected the designated object, the rotation detection unit determines the amount of rotation deviation of the plurality of object detection sensors based on the detection information of the designated object. The rotation deviation amount detection device according to any one of claims 1 to 4, which detects each of the detection sensors. The parameter used for detecting the object is changed depending on whether the rotation detection unit detects the rotation deviation amount of the object detection sensor or when the rotation detection unit does not detect the rotation deviation amount of the object detection sensor. 6. The instruction unit according to claim 1, further comprising: an instruction unit for instructing the object detection sensor of an operation mode and the parameter according to whether or not to detect a rotation deviation amount of the detection sensor. Rotational deviation amount detection device. The rotation deviation amount detection according to any one of claims 1 to 6, further comprising: a display control unit configured to arrange an image based on the detection information on a screen of a display unit in association with a position of the object detection sensor. apparatus. Further comprising a correction unit for correcting image distortion based on the detection information,
The rotation deviation amount detection device according to claim 7, wherein the display control unit displays the image in which the distortion is corrected on a screen of the display unit.   The rotation deviation amount detection device according to claim 8, wherein the display control unit arranges the images in which the distortions are corrected and displays the images on the screen of the display unit.   The rotation deviation amount detection device according to claim 9, wherein the display control unit combines the images in which the distortions are corrected and displays the combined images on the screen of the display unit. A sensor that detects an object,
An acquisition unit that acquires specific information that is information for specifying whether or not the specified object is detected by image recognition,
The parameters used to detect the object are changed and designated depending on whether the rotation deviation amount of the self-object detection sensor is detected by another device or when the rotation deviation amount of the self-object detection sensor is not detected by the other device. An image recognition unit that specifies whether or not the object is detected based on the specific information by image recognition,
A transmission unit that transmits the detection information indicating the result of detecting the object that has moved along the registered path and the identification information of the self-object detection sensor to the other device,
An object detection sensor including. A plurality of object detection sensors installed in a plurality of positions to detect an object,
A registration unit that registers the correct answer data regarding the path of the moving object and the path;
Based on the result of comparison between the correct information and the detection information representing the result of detection of the object moved along the registered path by the plurality of object detection sensors, the object detection sensor of the predetermined orientation A rotation detector that detects the amount of rotation deviation,
A rotation deviation amount detection system including. A plurality of object detection sensors for detecting an object are installed at a plurality of positions, a rotation deviation amount detection method executed by a rotation deviation amount detection device for detecting a rotation deviation amount of the object detection sensor,
Registering the correct answer data regarding the path of the moving object and the path,
The object detection sensor for a predetermined direction based on a comparison result of detection information indicating a result of detection of the object moving along the registered path by the plurality of object detection sensors and the correct answer data. Detecting the amount of rotation deviation of
A method for detecting the amount of rotation deviation including. A plurality of object detection sensors for detecting an object are installed at a plurality of positions, the computer of the rotation deviation amount detection device for detecting the rotation deviation amount of the object detection sensor,
A step of registering the correct answer data regarding the path of the moving object and the path;
Based on the result of comparison between the correct information and the detection information representing the result of detection of the object moved along the registered path by the plurality of object detection sensors, the object detection sensor of the predetermined orientation Procedure to detect the amount of rotation deviation,
A rotation deviation amount detection program for executing.

Images