CN106020364A - Smart bracelet with searching function of mute mobile phone and hand gesture recognition method - Google Patents

Abstract

The invention provides a smart bracelet with a searching function of a mute mobile phone. The smart bracelet comprises a power supply unit, a voltage stabilizing unit, a Bluetooth unit, an upper computer Bluetooth interface, a singlechip microcomputer unit and a signal acquisition acceleration sensor, wherein the power supply unit is electrically connected with the voltage stabilizing unit; the voltage stabilizing unit is respectively and electrically connected with the Bluetooth unit, the singlechip microcomputer unit and the signal acquisition acceleration sensor; the Bluetooth unit is respectively connected with the upper computer Bluetooth interface and the singlechip microcomputer unit; the singlechip microcomputer unit is electrically connected with the signal acquisition acceleration sensor; the power supply unit supplies power to the whole bracelet. The smart bracelet provided by the invention is reasonable in design and convenient to use; by adoption of a Bluetooth communication mode, the smart bracelet has the advantages of time saving, labor saving, convenience for users, small size, easiness for carrying, low power consumption and long endurance time; the smart bracelet can be wirelessly charged and is very convenient.

Description

A smart bracelet with a silent mobile phone search function and gesture recognition method

technical field

The invention relates to the technical field of electronic products, in particular to an intelligent bracelet with a silent mobile phone search function and a gesture recognition method.

Background technique

With the rapid development of science and technology, the smart wear industry is gradually emerging. Smart wearable devices have undoubtedly become one of the hottest topics at present. The advent of the era of wearable smart devices means the extension of human intelligence. Through these devices, people can better perceive external and their own information, and can process information more efficiently with the assistance of computers, networks and even other people. Enables more seamless communication.

In today's increasingly busy world, people spend most of the day at work. The mobile phone mute function is very useful for most people, especially for students and office workers. Living and working in this state, in order not to affect study and work, I had to turn my mobile phone to silent mode or even turn off vibration. For some careless friends, it is a common thing that the mobile phone cannot be found. Sometimes the mobile phone is lost, and sometimes the mobile phone is really lost, not lost. At this time, the phone can neither ring nor vibrate, and it is often distressed because the mobile phone cannot be found.

At present, although various mobile phone manufacturers have the function of retrieving mobile phones, such as Baidu Cloud, OPPO, Xiaomi, etc., most of them are for the retrieval of stolen mobile phones. Moreover, the operation process also requires a computer connected to the Internet. Logging in to a special website and entering the account password is not only very troublesome, but also cannot solve practical problems quickly and conveniently. There are also some that use App to find mobile phones, such as an interesting app called Whistle Android Finder. Whistle can be used to wake up the phone when the screen is off. Using a voice-based human-computer interaction method, the whistle can wake up the voice assistant of the mobile phone to complete the voice operation. You need to order a special whistle, your mobile phone needs a voice assistant, and you need to be connected to the Internet during use, otherwise it will not work.

Contents of the invention

The purpose of the present invention is to solve the above-mentioned deficiencies, providing a smart bracelet with a silent mobile phone search function and a gesture recognition method.

The purpose of the present invention is achieved through the following technical solutions:

A smart bracelet with the function of finding a silent mobile phone, including a power supply unit, a voltage stabilizing unit, a Bluetooth unit, a host computer Bluetooth interface, a single-chip microcomputer unit, and a signal acquisition acceleration sensor. The power supply unit is electrically connected to the voltage stabilizing unit, and the voltage stabilizing unit is electrically Connect the Bluetooth unit, single-chip unit and signal acquisition acceleration sensor, the Bluetooth unit is respectively connected to the upper computer Bluetooth interface and the single-chip unit, the single-chip unit is electrically connected to the signal acquisition acceleration sensor, and the power supply unit supplies power to the entire bracelet.

The signal acquisition acceleration sensor includes MPU6050 three-axis acceleration sensor.

The single-chip microcomputer unit includes the STC15F2K60S2 single-chip microcomputer chip.

The power supply unit is charged by wireless electromagnetic induction.

The present invention has following beneficial effect:

The invention is reasonable in design, easy to use, adopts bluetooth communication mode, saves time and effort, is convenient for users, small in size, easy to carry, low in power consumption, long in battery life, can choose to carry out wireless charging, is very convenient, and has the function of searching for mobile phones. In the silent mode or even off the vibration state, the mobile phone can ring and the wristband vibrate only through simple gestures, thereby reducing the distress of finding the mobile phone. As a result, the contact vibration can effectively wake you up from a deep sleep. The smart bracelet works with an APP developed based on an Android phone. A special feature of the APP is voltage collection. There is a power usage status in the upper left corner of the software interface, just like a mobile phone. Display the remaining battery power of the device and prompt the user to charge.

Description of drawings

Fig. 1 is the overall structure schematic diagram of the present invention;

Fig. 2 is a sensor circuit diagram of the present invention;

Fig. 3 is bluetooth unit circuit diagram of the present invention;

Fig. 4 is the circuit diagram of the voltage stabilizing unit of the present invention;

Fig. 5 is a circuit diagram of a power supply unit of the present invention;

Fig. 6 is the single-chip microcomputer unit circuit diagram of the present invention;

Fig. 7 is three components of human motion of the present invention;

Fig. 8 is a schematic diagram of gesture movement of the present invention;

Fig. 9 is a schematic diagram of complementary filtering of the present invention;

Fig. 10 is a schematic diagram of gesture recognition in the present invention;

Fig. 11 is a functional block diagram of the pedometer of the present invention;

Fig. 12 is a pedometer program block diagram of the present invention;

Fig. 13 is the mobile phone search program block diagram of the present invention;

Fig. 14 is the mobile phone incoming call program block diagram of the present invention;

Fig. 15 is a program block diagram of the mobile phone alarm clock of the present invention;

Fig. 16 is a program flow chart of the present invention;

Fig. 17 is real number of steps, monitoring steps and error graph of the present invention;

Fig. 18 is a test accuracy curve diagram of the present invention.

detailed description

Below in conjunction with accompanying drawing, the present invention will be further described:

As shown in Figure 1, a smart bracelet with a silent mobile phone search function includes a power supply unit 1, a voltage stabilization unit 2, a Bluetooth unit 3, a host computer Bluetooth interface 4, a single-chip microcomputer unit 5, and a signal acquisition acceleration sensor 6. Unit 1 is electrically connected to the voltage stabilizing unit 2, and the voltage stabilizing unit 2 is electrically connected to the Bluetooth unit 3, the single-chip microcomputer unit 5 and the signal acquisition acceleration sensor 6 respectively, and the Bluetooth unit 3 is respectively connected to the upper computer Bluetooth interface 4 and the single-chip microcomputer unit 5, and the single-chip microcomputer unit 5 is electrically connected to The signal is collected by the acceleration sensor 6, and the power supply unit 1 supplies power to the entire wristband.

The signal acquisition acceleration sensor includes MPU6050 three-axis acceleration sensor.

The single-chip microcomputer unit includes the STC15F2K60S2 single-chip microcomputer chip.

The power supply unit is charged by wireless electromagnetic induction.

Smart bracelet with STC15F2K60S2 microcontroller as processor. The smart bracelet uses the gravitational acceleration gyro sensor MPU-6050 (3-axis gyro sensor and 3-axis acceleration sensor), combined with the APP developed based on Android, and through smart Bluetooth communication, it realizes gesture control to mute mobile phone search, alarm clock status, step counting and Features like wireless charging. It has the characteristics of small size, low power consumption, stable and accurate work, and friendly human-computer interaction.

The smart bracelet is designed for busy office workers. The smart bracelet is worn on the main wrist, just like a watch. The present invention selects the gravitational acceleration gyroscope sensor MPU-6050 three-axis acceleration sensor to collect gesture action signals. When the acceleration sensor perceives the three-dimensional activity of the hand, the detected motion signal can directly enter the microprocessor through the I ² C communication interface. Then communicate with the host computer via Bluetooth. After signal preprocessing, signal resampling, feature extraction and gesture recognition, it is converted into motion information to drive related applications.

The system has functions such as mobile phone search, mobile phone incoming calls, and mobile phone alarm clock. Use the host computer software to process the collected data to realize storage and real-time display functions. The smart bracelet abandons the traditional buttons and display functions, all of which are displayed on the APP developed based on Android, and all controls are carried out through designed gestures.

Bluetooth communication principle: Android mobile phone turns on Bluetooth and establishes a connection with the Bluetooth serial port module of the wristband. The MCU sends data, and the Bluetooth serial port receives the serial port data and converts it into Bluetooth data. The Android mobile phone obtains instructions by reading the Bluetooth data, and then the mobile phone rings (play Ringtone) (MCU serial port→Bluetooth serial port→Android mobile phone). If the Bluetooth connection is not successful, the 24th pin (LED) will continuously output high and low pulses for 200ms. Using this feature, the red LED and green LED are used as display prompts. If the bluetooth is not connected, the red light keeps flashing, and the bluetooth connection is successful, the red light is off and the green light is on.

Linear voltage regulator circuit: The power input is 5V, and the system power supply only needs 3.7V, so a power regulator circuit is designed. Adopt XC6209 voltage regulator chip, after voltage stabilization, output 3.7V, suitable for 3.3V microcontroller.

Wireless charging uses electromagnetic induction, similar to a transformer. There is a coil at the sending end and a receiving end respectively. The coil at the sending end is connected to a wired power supply to generate an electromagnetic signal, and the receiving coil induces the electromagnetic signal at the sending end to generate current to charge the mobile phone. Charging time is no different from conventional chargers. The wireless charging transmitter chip is XKT-510 from Coretech, the wireless charging receiving chip is T3168, and the coil is 10uH.

The smart bracelet has the following functions: silent mobile phone search function, remote gesture control alarm clock, smart step counting function, battery life reminder, and wireless charging technology.

The six data (three-axis acceleration value and three-axis angular velocity value) of MPU6050 read through I2C will be converted into attitude at last. In order to reduce the calculation amount and facilitate the realization on STM32 main control, it can be converted into quaternion representation. Convert the six data into quaternions (q0, q1, q2, q3), and then convert the quaternions into Euler angles (Pitch, Roll, Yaw angles). Pitch and Roll angles can be obtained through both acceleration and angular velocity (acceleration cannot obtain Yaw angles), that is to say, there are two sets of Pitch and Roll angles.

The accelerometer and gyroscope of the MPU6050 have their own advantages and disadvantages. The acceleration value of the three axes has long-term stability characteristics, but the response speed is slow and there is no cumulative error. The inclination angle can be obtained through calculation, but it contains too much noise and cannot be used directly; The impact of external vibration is small, the real-time performance is very good, and the accuracy is high. The inclination angle can be obtained by integrating the angular velocity, but the cumulative error will be integrated due to the existence of zero drift, temperature drift, nonlinear and other factors, which cannot be eliminated by itself. Therefore, the accelerometer or gyroscope of the MPU6050 cannot be used alone to obtain the inclination angle, and it needs to be complementary. Angle corrections are to be made via the accelerometer. The gyroscope should be high-passed to suppress low-frequency drift, and the accelerometer to be low-passed to suppress noise, and the two should be fused in a certain ratio to obtain a more accurate reflection of the change in angle.

Complementary filtering using the quaternion method. Complementary filtering algorithm has a small amount of calculation and works well in the case of weak coupling. After complementary filtering and fusion, the fused data is calculated by inclination to obtain the angle, and finally the gesture is judged through the time window.

Acquire gesture acceleration signals, perform motion window detection, automatically locate possible gesture motion sequences, and then eliminate the influence of gesture motion amplitude and speed differences through smoothing filtering and signal window resampling. Finally, the gesture action recognition is converted into action information to drive the phone to ring. The time window: minimum 15ms, maximum 85ms, requires 5 trainings to accurately recognize, and the sampling time is 50ms.

The movement of the human body includes three components, which are forward, vertical and lateral. As shown in Figure 1, the three axes (x-axis, y-axis, z-axis) of the sensor correspond to these three directions, which represents Acceleration values in three directions: vertical, forward and lateral when the human body is walking. In actual application, the sensor is in an unknown direction, so the three-axis measurement value of the sensor does not correspond to the three components of motion, but the projection value of the motion acceleration on the three axes of the sensor.

Generally speaking, the human body walks 0.5-2.0 steps per second, no more than 5.0 steps. So a reasonable pedometer output would be 0.5-5.0Hz. A FIR low-pass filter with a cutoff frequency of 5.0 Hz is used to filter out high-frequency noise.

The modulus of the three acceleration signals output by the acceleration sensor is summed to obtain the acceleration model of the human body movement. Since the data will be interfered by the sensor itself and various external noises, the original data needs to be filtered before waveform analysis. After the signal passes through the low-pass filter to filter out the signal with a frequency above 5.0Hz, then the dynamic threshold is judged through the adaptive threshold, and then the number of human motions is detected.

For sports, the number of steps is a quantifiable indicator. When the number of steps is used to represent human movement, the accuracy of step detection is a measure of the pedometer system, because the higher the accuracy of the pedometer, the higher the detection accuracy. The more accurate the number of exercise steps is, the more it can reflect the human body's exercise situation, exercise volume and exercise activity. The definition of step counting accuracy is shown in formula 1:

Table 1 shows the test results of step count detection performed on different testees under different motion patterns.

Through the analysis of test results, the accuracy of slow walking and fast walking reaches about 97%. Running accuracy is slightly down, but also greater than 90%. The step counting accuracy predetermined by the system is 95%, and reaching this accuracy meets the system requirements.

Use the pedometer to detect the number of steps of different subjects under different exercise forms. The test results show that the step counting accuracy of the wrist-type pedometer is higher than 95%, which meets the design requirements.

The invention has the function of searching for the mobile phone. When the mobile phone is set to the silent mode or even the vibration is turned off, the mobile phone can ring and the bracelet vibrate only through simple gestures, thereby reducing the trouble of finding the mobile phone. The remote gesture control alarm clock becomes It is more user-friendly, and wakes up with a vibration effect at the same time. The contact vibration can effectively wake you up from a deep sleep. The smart bracelet works with the APP developed based on Android. A special feature of the APP is voltage collection. There is a Power usage, just like a mobile phone, displays the remaining battery power of the device and prompts the user to charge.

Table 1 Test results

tester real steps Monitor steps error Accuracy walk slowly User A 100 101 +1 99% User B 100 99 -1 99% User C 100 102 +2 98% go quickly User A 100 105 +5 95% User B 100 103 +3 97% User C 100 104 +4 96% running User A 100 109 +9 91% User B 100 110 +10 90% User C 100 105 +5 95%

Claims (6)

1. An intelligent wristband and a gesture recognition method with a silent mobile phone search function, comprising a power supply unit, a voltage stabilizing unit, a bluetooth unit, a host computer bluetooth interface, a single-chip unit and a signal acquisition acceleration sensor, characterized in that: the power supply unit Connect the voltage stabilizing unit, the voltage stabilizing unit is electrically connected to the Bluetooth unit, the single-chip microcomputer unit and the signal acquisition acceleration sensor respectively, the Bluetooth unit is respectively connected to the upper computer Bluetooth interface and the single-chip microcomputer unit, the single-chip microcomputer unit is electrically connected to the signal acquisition acceleration sensor, and the power supply unit supplies power to the whole bracelet . 2. A smart wristband with a silent mobile phone search function according to claim 1, characterized in that: said signal acquisition acceleration sensor includes an MPU6050 triaxial acceleration sensor. 3. A kind of smart wristband with the function of finding a silent mobile phone according to claim 1, characterized in that: the single-chip microcomputer unit comprises a STC15F2K60S2 single-chip microcomputer chip. 4. The smart bracelet with the function of finding a silent mobile phone according to claim 1, wherein the power supply unit is charged by wireless electromagnetic induction. 5. A gesture recognition method with a silent mobile phone search function smart bracelet, characterized in that: the algorithm is as follows: read six data (three-axis acceleration value, three-axis angular velocity value) of MPU6050 by I2C to be converted into attitude at last , in order to reduce the amount of calculation and facilitate implementation on the STM32 master, it can be converted into a quaternion representation; convert the six data into a quaternion (q0, q1, q2, q3), and then convert the quaternion into Euler angles (Pitch, Roll, Yaw corners). Pitch and Roll angles can be obtained through both acceleration and angular velocity (acceleration cannot obtain Yaw angles), that is to say, there are two sets of Pitch and Roll angles. Using the quaternion method complementary filtering, after complementary filtering fusion, the fused data is calculated through the inclination angle to obtain the angle, and finally the gesture is judged through the time window, the gesture acceleration signal is collected, the action window is detected, and the possible gesture action sequence is automatically located, and then Through smoothing filtering and signal window resampling, the influence of gesture range and speed difference is eliminated, and finally the gesture recognition is converted into motion information to drive the phone to ring. 6. A gesture recognition method with a silent mobile phone search function smart bracelet, characterized in that: it also includes a step counting method: the motion of the human body includes three components, which are forward, vertical and sideways, and the three components of the sensor The axes (x-axis, y-axis, z-axis) exactly correspond to these three directions, which represent the acceleration values in the vertical, forward and lateral directions of the human body when walking, and the modulus of the three-way acceleration signals output by the acceleration sensor Sum up to get the acceleration model of human body movement. Since the data will be interfered by the sensor itself and various external noises, the original data needs to be filtered before waveform analysis, so that the signal can be filtered by a low-pass filter to filter frequencies above 5.0Hz After receiving the signal, the dynamic threshold judgment is performed through the adaptive threshold, and then the number of human motions is detected.