WO2020207071A1 - 一种健身动作识别方法、***及电子设备 - Google Patents

一种健身动作识别方法、***及电子设备 Download PDF

Info

Publication number
WO2020207071A1
WO2020207071A1 PCT/CN2019/130588 CN2019130588W WO2020207071A1 WO 2020207071 A1 WO2020207071 A1 WO 2020207071A1 CN 2019130588 W CN2019130588 W CN 2019130588W WO 2020207071 A1 WO2020207071 A1 WO 2020207071A1
Authority
WO
WIPO (PCT)
Prior art keywords
heart rate
data
motion
fitness
axis
Prior art date
Application number
PCT/CN2019/130588
Other languages
English (en)
French (fr)
Inventor
赵国如
郭贵昌
宁运琨
李慧奇
王成
黄连鹤
Original Assignee
深圳先进技术研究院
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳先进技术研究院 filed Critical 深圳先进技术研究院
Publication of WO2020207071A1 publication Critical patent/WO2020207071A1/zh
Priority to US17/481,323 priority Critical patent/US20220001262A1/en

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B71/00—Games or sports accessories not covered in groups A63B1/00 - A63B69/00
    • A63B71/06—Indicating or scoring devices for games or players, or for other sports activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B71/00—Games or sports accessories not covered in groups A63B1/00 - A63B69/00
    • A63B71/06—Indicating or scoring devices for games or players, or for other sports activities
    • A63B71/0686—Timers, rhythm indicators or pacing apparatus using electric or electronic means
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/20—Movements or behaviour, e.g. gesture recognition
    • G06V40/23—Recognition of whole body movements, e.g. for sport training
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/17—Counting, e.g. counting periodical movements, revolutions or cycles, or including further data processing to determine distances or speed
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/30—Speed
    • A63B2220/34—Angular speed
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/40—Acceleration
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/62—Time or time measurement used for time reference, time stamp, master time or clock signal
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/80—Special sensors, transducers or devices therefor
    • A63B2220/803—Motion sensors
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/80—Special sensors, transducers or devices therefor
    • A63B2220/83—Special sensors, transducers or devices therefor characterised by the position of the sensor
    • A63B2220/836—Sensors arranged on the body of the user
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2230/00—Measuring physiological parameters of the user
    • A63B2230/04—Measuring physiological parameters of the user heartbeat characteristics, e.g. ECG, blood pressure modulations
    • A63B2230/06—Measuring physiological parameters of the user heartbeat characteristics, e.g. ECG, blood pressure modulations heartbeat rate only
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching
    • G06F2218/14—Classification; Matching by matching peak patterns

Definitions

  • This application belongs to the technical field of exercise state recognition, and in particular relates to a fitness action recognition method, system and electronic equipment.
  • the recognition of the motion state can be divided into the following two directions:
  • Motion state recognition based on image and video This method mainly captures the motion category of the human body by analyzing the data collected by the mining camera. Since the data collected by the camera is easily affected by factors such as weather, light, distance, orientation, etc., the scenes used are also very limited, and the video images take up storage space and cannot be used for a long time.
  • This method mainly collects data from the sensor in the wearable device carried and then analyzes and researches it. Compared with the motion state recognition method based on image and video, this method has the following advantages: a. Low cost and easy to carry: the wearable device is cheap and small and can be worn with you; b. Strong anti-interference: the data collection process is affected by the external environment The impact is small; c. Ability to continuously obtain data: Carrying it with you can ensure continuous data acquisition.
  • the existing motion state recognition based on wearable devices are all based on inertial sensors to collect motion data, so the judgment of the motion state is limited, and it is impossible to accurately distinguish fast running and jogging, and the current motion state recognition is for the daily life of the human body. Activities, such as walking, running, standing up, sitting down, etc., cannot be identified for the fitness crowd.
  • Chinese patent 201410306132.7 discloses a human body motion analysis method and device based on heart rate and acceleration sensors.
  • the device can detect unobvious exercise states such as weight lifting, strength training, and yoga.
  • the patent is based on the human body motion analysis method of heart rate and acceleration sensors, which can effectively detect various aerobic and anaerobic exercises and sleep, and prevent misjudgments caused by waving hands and folding quilts.
  • the patent only uses these data to distinguish whether the human body is in a motion state or a non-motion state, and cannot determine what the human body is doing, and it cannot effectively reflect the human body's motion status.
  • This application provides a method, system and electronic device for recognizing fitness actions, which aim to solve one of the above technical problems in the prior art at least to a certain extent.
  • a method for recognizing fitness actions includes the following steps:
  • Step a Collect the motion data and heart rate data of the human body during movement through the nine-axis inertial sensor and the heart rate sensor;
  • Step b using a motion recognition algorithm to calculate the combined acceleration, combined angular velocity, roll angle, and real-time heart rate values of the nine-axis inertial sensor based on the motion data and heart rate data;
  • Step c Recognizing fitness actions based on the characteristics of the combined acceleration, combined angular velocity and roll angle of the nine-axis inertial sensor and the real-time heart rate value.
  • the technical solution adopted in the embodiment of the application further includes: in the step a, the motion recognition algorithm is used to calculate the combined acceleration, the combined angular velocity, and the roll of the nine-axis inertial sensor according to the motion data and heart rate data.
  • the angle and the real-time heart rate value specifically include: filtering the collected heart rate data to remove motion artifacts to obtain a real-time heart rate value.
  • the real-time heart rate value includes the maximum exercise heart rate, the minimum exercise heart rate, and the resting heart rate.
  • the technical solution adopted in the embodiment of the application further includes: in the step a, the motion recognition algorithm is used to calculate the combined acceleration, the combined angular velocity, and the roll of the nine-axis inertial sensor according to the motion data and heart rate data.
  • Angular and real-time heart rate values also include: calibrating and filtering the collected exercise data to obtain three-axis acceleration, three-axis angular velocity, and three-axis magnetometer data; performing three-axis acceleration, three-axis angular velocity, and three-axis magnetometer data Fusion to obtain the quaternion required for the final acceleration, the final angular velocity and the attitude calculation.
  • the technical solution adopted in the embodiment of the application further includes: in the step a, the motion recognition algorithm is used to calculate the combined acceleration, the combined angular velocity, and the roll of the nine-axis inertial sensor according to the motion data and heart rate data.
  • the angle and real-time heart rate value also include: fusing the three-axis acceleration, three-axis angular velocity, and three-axis magnetometer data to obtain the combined acceleration, combined angular velocity, and quaternion required for attitude calculation; Convert data to obtain the attitude angle, roll angle and heading angle data.
  • the technical solution adopted in the embodiment of the present application further includes: after step c, it further includes: timing or counting the fitness action according to the fitness action recognition result, and performing a reminder operation according to the set time threshold or frequency threshold.
  • a fitness action recognition system including:
  • Inertial sensor module used to collect the motion data of the human body through the nine-axis inertial sensor;
  • Heart rate sensor module used to collect heart rate data when the human body moves through the heart rate sensor
  • Motion recognition algorithm module used to use motion recognition algorithms to calculate the combined acceleration, combined angular velocity, roll angle, and real-time heart rate values of the nine-axis inertial sensor based on the motion data and heart rate data;
  • Fitness action recognition module used for recognizing fitness actions based on the characteristics of the combined acceleration, combined angular velocity, and roll angle of the nine-axis inertial sensor and the real-time heart rate value.
  • the motion recognition algorithm module further includes:
  • Heart rate data processing unit used to filter the collected heart rate data, remove motion artifacts, and obtain real-time heart rate values.
  • the real-time heart rate values include the maximum exercise heart rate, the minimum exercise heart rate, and the resting heart rate.
  • the motion recognition algorithm module further includes:
  • Motion data processing unit used to calibrate and filter the collected motion data to obtain three-axis acceleration, three-axis angular velocity and three-axis magnetometer data;
  • Data fusion unit used to fuse three-axis acceleration, three-axis angular velocity, and three-axis magnetometer data to obtain the combined acceleration, combined angular velocity, and quaternion required for attitude calculation.
  • the motion recognition algorithm module further includes:
  • Data conversion unit used to convert the quaternion to obtain attitude angle, roll angle and heading angle data respectively.
  • Fitness reminder module used to time or count fitness actions according to the recognition results of the fitness actions, and perform reminding operations according to the set time threshold or frequency threshold.
  • an electronic device including:
  • At least one processor At least one processor
  • a memory communicatively connected with the at least one processor; wherein,
  • the memory stores instructions that can be executed by the one processor, and the instructions are executed by the at least one processor, so that the at least one processor can perform the following operations of the aforementioned fitness action recognition method:
  • Step a Collect the motion data and heart rate data of the human body during movement through the nine-axis inertial sensor and the heart rate sensor;
  • Step b using a motion recognition algorithm to calculate the combined acceleration, combined angular velocity, roll angle, and real-time heart rate values of the nine-axis inertial sensor based on the motion data and heart rate data;
  • Step c Recognizing fitness actions based on the characteristics of the combined acceleration, combined angular velocity, and roll angle of the nine-axis inertial sensor and the real-time heart rate value.
  • the beneficial effects produced by the embodiments of the present application are: the fitness action recognition method, system and electronic equipment of the embodiments of the present application collect exercise data and heart rate by wearing nine-axis inertial sensors and heart rate sensors on the person’s body Data, through the exercise data and heart rate data design exercise state recognition algorithm, through the real-time data collection, the processor uses the exercise recognition algorithm to identify fitness actions based on the characteristics of the exercise data and real-time heart rate data, and clearly recognizes fast running And jogging can improve the fitness efficiency of the fitness crowd, and guide the training of the fitness crowd better and more conveniently.
  • Fig. 1 is a flowchart of a fitness action recognition method according to an embodiment of the present application
  • Figure 2 is a schematic diagram of the characteristics of poppy jump action
  • Figure 3 is a schematic diagram of the characteristics of the pull-up movement
  • Figure 4 is a schematic diagram of squat action characteristics
  • Figure 5 is a schematic diagram of the action characteristics of sit-ups
  • Figure 6 is a schematic diagram of the characteristics of the high leg raise
  • Figure 7 is a schematic diagram of the characteristics of the opening and closing jump action
  • Figure 8 is a schematic diagram of the deadlift action characteristics
  • Figure 9 is a schematic diagram of running action features
  • FIG. 10 is a hardware system frame diagram of a fitness action recognition system according to an embodiment of the application.
  • FIG. 11 is a schematic structural diagram of a fitness action recognition system according to an embodiment of the present application.
  • FIG. 12 is a schematic diagram of the hardware device structure of the fitness action recognition method provided by an embodiment of the present application.
  • FIG. 1 is a flowchart of a fitness action recognition method according to an embodiment of the present application.
  • the fitness action recognition method of the embodiment of the present application includes the following steps:
  • Step 100 Collect the motion data (acceleration, angular velocity, magnetic intensity, etc.) and heart rate data of the human body during motion through the nine-axis inertial sensor and the heart rate sensor;
  • step 100 the motion data collection is completed by STM32 and MPU9250.
  • STM32 and MPU9250 are connected through the IIC bus.
  • the MCU is set through the corresponding registers of MPU9250, including registers such as sampling rate and sensor range.
  • the default acceleration range It is ⁇ 8g
  • the gyroscope is ⁇ 1000dbps
  • the magnetometer works in single measurement mode, which can be set according to actual operation.
  • Each sensor can output 6 bytes of data once sampling, and the output of each sensor's three axes occupies 2 bytes, with the high order first.
  • the heart rate data collection is completed by STM32 and the heart rate sensor.
  • the heart rate sensor is connected to the STM32 through the IIC bus, and the configuration of its registers is performed.
  • Step 110 Filtering the collected heart rate data to remove motion artifacts to obtain a real-time heart rate value
  • step 110 the heart rate calculation method is:
  • Normal resting heart rate is usually 60-100 beats/min for adults with normal resting heart rate.
  • h i When the human body is at rest, according to the heart rate sensor data, record it every 10 seconds (h i ), record 5 groups continuously, find the average, and then multiply by 6. Get the resting heart rate per minute (heart):
  • Step 120 Perform data calibration and filter processing on the collected motion data to obtain three-axis acceleration, three-axis angular velocity, and three-axis magnetometer data;
  • Step 130 Fuse the data of the three-axis acceleration, the three-axis angular velocity and the three-axis magnetometer to obtain the combined acceleration, the combined angular velocity and the quaternion required for the attitude calculation;
  • step 130 the purpose of data fusion is to obtain the quaternion required for the attitude calculation.
  • the quaternion calculation is small, has no singularities, and can meet the real-time calculation of the attitude of the aircraft during movement.
  • the size and direction they represent must be the same, but due to the error in the rotation matrix of the two coordinate systems, when a vector passes through the rotation matrix with error In another coordinate system, there will be a deviation from the theoretical value.
  • the system can correct the rotation matrix through this deviation.
  • the element of the rotation matrix is a quaternion, and the corrected quaternion can be converted into an attitude angle with a smaller error. .
  • Step 140 Convert the quaternion to obtain the attitude angle Pitch (pitch angle), Roll (roll angle) and Yaw (heading angle) data respectively.
  • Step 150 Recognize fitness actions based on the characteristics of the combined acceleration, the combined angular velocity and the roll angle (Roll) and the real-time heart rate value;
  • step 150 the characteristics of the combined acceleration, the combined angular velocity and the roll angle (Roll) and the heart rate value corresponding to each fitness action are different.
  • the following are respectively pop-ups, pull-ups, squats, sit-ups,
  • the actions of raising legs, opening and closing jumping, deadlifting, and running (fast running and jogging) are explained in detail.
  • Figures 2-9 they are schematic diagrams of the action characteristics of poppy jump, pull-ups, squats, sit-ups, high leg lifts, open and close jumps, deadlifts, and running (fast running and jogging).
  • Step 160 Perform corresponding timing/counting according to the fitness action recognition result, and perform a reminder operation according to the set time/time threshold.
  • step 160 take poppy jump, pull-ups, squats, sit-ups, leg lifts, open and close jumps, deadlifts, and running actions as examples.
  • the fitness action recognition results are poppy jump, open and close jump, Perform timing when raising your legs or running, and remind once when the timing reaches the set timing threshold (in the embodiment of this application, the timing threshold is set to one minute, which can be specifically set according to actual operations);
  • the result is that deadlifts, pull-ups, squats or sit-ups are counted, and the count reaches the set count threshold (in the embodiment of this application, the count threshold is set to 10 times, which can be set according to actual operations. Remind once at a certain time.
  • FIG. 10 is a hardware system framework diagram of the fitness action recognition system according to an embodiment of the application.
  • the hardware system includes inertial sensor module, heart rate sensor module, USB conversion module, firmware download interface, USB power supply interface and main control module.
  • the main control module adopts STM32F407ZGT6 chip, its main frequency is up to 168MHZ, 1MB FLASH, 192KB SRAM provides fast operation and processing capabilities for reliable and stable wireless sensor network programs and high-speed real-time storage of data.
  • LQFP144 ultra-small package Realize the miniaturization of the entire sensor node.
  • the chip of the USB conversion module is CP2102, and its communication protocol with the main control module is USART. It has the characteristics of high integration. It can be built-in USB2.0 full-speed function controller, USB transceiver, crystal oscillator, EEPROM and asynchronous serial data
  • the bus (UART) supports the full-function signal of the modem without any external USB devices. It can complete the level conversion and communication control of the RS232 protocol and USB2.0 protocol of the USART interface of the sensor network node.
  • the inertial sensor module As the data source of the system, the inertial sensor module, IMU (Inertial Measurement Unit) needs to have high reliability, high stability and anti-interference ability.
  • MPU9250 integrates 3-axis acceleration, 3-axis gyroscope and digital motion processor (DMP), and can directly output all 9-axis data via SPI or I2C. The range of nine-axis data is programmable.
  • the chip is packaged in QFN, which is conducive to reducing the volume of the entire system. Multi-range options can meet the system's requirements for collecting various human body movement data.
  • DMP provides a variety of data fusion methods for it; low power consumption mode can be used in static Timely reduce system power consumption and meet system requirements for low power consumption.
  • FIG. 11 is a schematic structural diagram of a fitness action recognition system according to an embodiment of the present application.
  • the fitness action recognition system of the embodiment of the application includes an inertial sensor module, a heart rate sensor module, a motion recognition algorithm module, a fitness action recognition module, and a fitness reminder module.
  • Inertial sensor module used to collect the motion data (acceleration, angular velocity, magnetic intensity, etc.) of the human body through the nine-axis inertial sensor; among them, the motion data collection is completed by STM32 and MPU9250, STM32 and MPU9250 are connected through IIC bus, MCU Set through the corresponding registers of the MPU9250, including the sampling rate, sensor range and other registers.
  • the default acceleration range is ⁇ 8g
  • the gyroscope is ⁇ 1000dbps
  • the magnetometer works in single measurement mode. The specific operation can be based on actual operation. Make settings.
  • Each sensor can output 6 bytes of data once sampling, and the output of each sensor's three axes occupies 2 bytes, with the high order first.
  • Heart rate sensor module used to collect the heart rate data of the human body through the heart rate sensor; among them, the heart rate data collection is completed by the STM32 and the heart rate sensor, the heart rate sensor is connected to the STM32 through the IIC bus, and its register configuration is performed.
  • Motion recognition algorithm module used to use motion recognition algorithm to calculate the combined acceleration, combined angular velocity, roll angle and real-time heart rate value of the nine-axis inertial sensor based on the motion data and heart rate data; specifically, the motion recognition algorithm module includes:
  • Heart rate data processing unit used to filter the collected heart rate data, remove motion artifacts, and obtain real-time heart rate values; among them, the heart rate value calculation method is:
  • Normal resting heart rate is usually 60-100 beats/min for adults with normal resting heart rate.
  • h i When the human body is at rest, according to the heart rate sensor data, record it every 10 seconds (h i ), record 5 groups continuously, find the average, and then multiply by 6. Get the resting heart rate per minute (heart):
  • Motion data processing unit used to calibrate and filter the collected motion data to obtain three-axis acceleration, three-axis angular velocity and three-axis magnetometer data;
  • Data fusion unit used to fuse three-axis acceleration, three-axis angular velocity and three-axis magnetometer data to obtain the combined acceleration, combined angular velocity and the quaternion required for attitude calculation; among them, the purpose of data fusion is to obtain the attitude
  • the quaternion required for the calculation, the calculation of the quaternion is small, there is no singularity, and it can meet the real-time calculation of the attitude of the aircraft during the movement.
  • the size and direction they represent must be the same, but due to the error in the rotation matrix of the two coordinate systems, when a vector passes through the rotation matrix with error In another coordinate system, there will be a deviation from the theoretical value.
  • the system can correct the rotation matrix through this deviation.
  • the element of the rotation matrix is a quaternion, and the corrected quaternion can be converted into an attitude angle with a smaller error. .
  • Data conversion unit used to convert the quaternion to obtain the attitude angle Pitch (pitch angle), Roll (roll angle) and Yaw (heading angle) data.
  • Fitness action recognition module It is used to recognize fitness actions through the characteristics of total acceleration, total angular velocity and roll angle (Roll) and real-time heart rate value; among them, each fitness action corresponds to the total acceleration, total angular velocity and roll angle ( Roll) features and heart rate values are different.
  • the following are performed with poppy jump, pull-ups, squats, sit-ups, high leg lifts, open and close jumps, deadlifts, and running (fast running and jogging) actions.
  • Specific instructions Specifically, as shown in Figures 2-9, they are schematic diagrams of the action characteristics of poppy jump, pull-ups, squats, sit-ups, leg lifts, open and close jumps, deadlifts and running.
  • Fitness reminder module used to perform corresponding timing/counting according to the result of fitness action recognition, and remind operation according to the set time/time threshold. Among them, take poppy jump, pull-ups, squats, sit-ups, high leg lifts, open and close jumps, deadlifts, and running actions as examples.
  • the fitness action recognition results are poppy jump, open and close jump, high lift Time the leg or running, and remind once when the time reaches the set timing threshold (in the embodiment of the present application, the timing threshold is set to one minute, which can be specifically set according to actual operation); the recognition result of the fitness action is Count during deadlifts, pull-ups, squats or sit-ups, and count when the count reaches the set count threshold (in the embodiment of this application, the count threshold is set to 10 times, which can be set according to actual operations) Remind once.
  • FIG. 12 is a schematic diagram of the hardware device structure of the fitness action recognition method provided by an embodiment of the present application.
  • the device includes one or more processors and memory. Taking a processor as an example, the device may also include: an input system and an output system.
  • the processor, the memory, the input system, and the output system may be connected through a bus or in other ways.
  • the connection through a bus is taken as an example in FIG. 12.
  • the memory can be used to store non-transitory software programs, non-transitory computer executable programs, and modules.
  • the processor executes various functional applications and data processing of the electronic device by running non-transitory software programs, instructions, and modules stored in the memory, that is, realizing the processing methods of the foregoing method embodiments.
  • the memory may include a program storage area and a data storage area, where the program storage area can store an operating system and an application program required by at least one function; the data storage area can store data and the like.
  • the memory may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid state storage devices.
  • the storage may optionally include storage remotely arranged with respect to the processor, and these remote storages may be connected to the processing system through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.
  • the input system can receive input digital or character information, and generate signal input.
  • the output system may include display devices such as a display screen.
  • the one or more modules are stored in the memory, and when executed by the one or more processors, the following operations of any of the foregoing method embodiments are performed:
  • Step a Collect the motion data and heart rate data of the human body during movement through the nine-axis inertial sensor and the heart rate sensor;
  • Step b using a motion recognition algorithm to calculate the combined acceleration, combined angular velocity, roll angle, and real-time heart rate values of the nine-axis inertial sensor based on the motion data and heart rate data;
  • Step c Recognizing fitness actions based on the characteristics of the combined acceleration, combined angular velocity and roll angle of the nine-axis inertial sensor and the real-time heart rate value.
  • the embodiment of the present application provides a non-transitory (nonvolatile) computer storage medium, the computer storage medium stores computer executable instructions, and the computer executable instructions can perform the following operations:
  • Step a Collect the motion data and heart rate data of the human body during movement through the nine-axis inertial sensor and the heart rate sensor;
  • Step b using a motion recognition algorithm to calculate the combined acceleration, combined angular velocity, roll angle, and real-time heart rate values of the nine-axis inertial sensor based on the motion data and heart rate data;
  • Step c Recognizing fitness actions based on the characteristics of the combined acceleration, combined angular velocity and roll angle of the nine-axis inertial sensor and the real-time heart rate value.
  • the embodiment of the present application provides a computer program product, the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program includes program instructions, when the program instructions are executed by a computer To make the computer do the following:
  • Step a Collect the motion data and heart rate data of the human body during movement through the nine-axis inertial sensor and the heart rate sensor;
  • Step b using a motion recognition algorithm to calculate the combined acceleration, combined angular velocity, roll angle, and real-time heart rate values of the nine-axis inertial sensor based on the motion data and heart rate data;
  • Step c Recognizing fitness actions based on the characteristics of the combined acceleration, combined angular velocity and roll angle of the nine-axis inertial sensor and the real-time heart rate value.
  • the fitness action recognition method, system, and electronic device of the embodiments of the application collect exercise data and heart rate data by wearing nine-axis inertial sensors and heart rate sensors on a person’s body, and design exercise state recognition algorithms through the exercise data and heart rate data.
  • the processor uses the exercise recognition algorithm to recognize fitness actions based on the characteristics of the exercise data and real-time heart rate data, and clearly recognizes fast running and jogging, which can improve the fitness efficiency of the fitness crowd, which is better and more convenient To guide the training of the fitness crowd.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Physical Education & Sports Medicine (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Psychiatry (AREA)
  • Social Psychology (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

一种健身动作识别方法、***及电子设备。所述方法包括：步骤a：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数据和心率数据；步骤b：利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；步骤c：根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。根据运动数据的特征和实时心率数据对健身动作进行识别，并很清楚的识别出快跑和慢跑，可以提高健身人群的健身效率，更好、更方便的指导健身人群的训练。

Description

一种健身动作识别方法、***及电子设备 技术领域
本申请属于运动状态识别技术领域，特别涉及一种健身动作识别方法、***及电子设备。
背景技术
目前，根据研究的数据类型不同可以将运动状态识别分成以下2个方向：
1)基于图像视频的运动状态识别：该方法主要通过分析挖掘摄像头采集的数据来捕捉人体的运动类别。由于摄像头采集数据很容易受天气、光线、距离、方位等因素的影响，使用的场景也非常有限，并且由于视频图像非常占用存储空间无法长期投入使用。
2)基于可穿戴设备的运动状态识别：该方法主要通过随身携带的穿戴设备中的传感器采集数据然后分析研究。相对于基于图像视频的运动状态识别方法，本方法具有以下几种优势：a、成本低且携带方便：穿戴设备价格低廉且小巧可以随身佩带；b、抗干扰性强：采集数据过程受外界环境影响小；c、持续获取数据的能力：随身携带可以保证持续地获取数据。
然而，现有的基于可穿戴设备的运动状态识别都是基于惯性传感器采集运动数据，因此判断运动状态有限，不能够准确的区分快跑、慢跑，并且现在的运动状态识别都是针对人体的日常活动，比如走路、跑步、起立、坐下等，而不能针对健身人群进行动作识别。
中国专利201410306132.7公开了一种基于心率和加速度传感器的人体运动分析方法及其装置。该装置能够检测到举重、力量训练、瑜伽等肢体动作不明 显的运动状态。该专利基于心率和加速度传感器的人体运动分析方法，能够进行有效的检测各种有氧运动和无氧运动以及睡眠，防止因挥手、叠被子造成误判。但是，该专利只是利用这些数据来区分人体是在运动状态还是在非运动状态，并不能判断人体具体是在做什么运动，不能有效的反映人体的运动状况。
发明内容
本申请提供了一种健身动作识别方法、***及电子设备，旨在至少在一定程度上解决现有技术中的上述技术问题之一。
为了解决上述问题，本申请提供了如下技术方案：
一种健身动作识别方法，包括以下步骤：
步骤a：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数据和心率数据；
步骤b：利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
步骤c：根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。
本申请实施例采取的技术方案还包括：在所述步骤a中，所述利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值具体包括：对采集的心率数据进行滤波处理，去除运动伪迹，得到实时心率值，所述实时心率值包括最大运动心率、最小运动心率和静息心率。
本申请实施例采取的技术方案还包括：在所述步骤a中，所述利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速 度、合角速度、横滚角以及实时心率值还包括：对采集的运动数据进行数据校准和滤波处理，得到三轴加速度、三轴角速度和三轴磁力计数据；将三轴加速度、三轴角速度和三轴磁力计数据进行融合，得到合加速度、合角速度以及姿态解算所需要的四元数。
本申请实施例采取的技术方案还包括：在所述步骤a中，所述利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值还包括：将所述三轴加速度、三轴角速度和三轴磁力计数据进行融合，得到合加速度、合角速度以及姿态解算所需要的四元数；并对所述四元数进行转换，分别得到姿态角、横滚角和航向角数据。
本申请实施例采取的技术方案还包括：所述步骤c后还包括：根据所述健身动作识别结果对健身动作进行计时或计数，并根据设定的时间阈值或次数阈值进行提醒操作。
本申请实施例采取的另一技术方案为：一种健身动作识别***，包括：
惯性传感器模块：用于通过九轴惯性传感器采集人体运动时的运动数据；
心率传感器模块：用于通过心率传感器采集人体运动时的心率数据；
运动识别算法模块：用于利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
健身动作识别模块：用于根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。
本申请实施例采取的技术方案还包括：所述运动识别算法模块还包括：
心率数据处理单元：用于对采集的心率数据进行滤波处理，去除运动伪迹，得到实时心率值，所述实时心率值包括最大运动心率、最小运动心率和静息心 率。
本申请实施例采取的技术方案还包括：所述运动识别算法模块还包括：
运动数据处理单元：用于对采集的运动数据进行数据校准和滤波处理，得到三轴加速度、三轴角速度和三轴磁力计数据；
数据融合单元：用于将三轴加速度、三轴角速度和三轴磁力计数据进行融合，得到合加速度、合角速度以及姿态解算所需要的四元数。
本申请实施例采取的技术方案还包括：所述运动识别算法模块还包括：
数据转换单元：用于对所述四元数进行转换，分别得到姿态角、横滚角和航向角数据。
本申请实施例采取的技术方案还包括：
健身提醒模块：用于根据所述健身动作识别结果对健身动作进行计时或计数，并根据设定的时间阈值或次数阈值进行提醒操作。
本申请实施例采取的又一技术方案为：一种电子设备，包括：
至少一个处理器；以及
与所述至少一个处理器通信连接的存储器；其中，
所述存储器存储有可被所述一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行上述的健身动作识别方法的以下操作：
步骤a：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数据和心率数据；
步骤b：利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
步骤c：根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征 以及实时心率值对健身动作进行识别。
相对于现有技术，本申请实施例产生的有益效果在于：本申请实施例的健身动作识别方法、***及电子设备通过在人的身上佩戴九轴惯性传感器和心率传感器等设备收集运动数据和心率数据，通过运动数据和心率数据设计运动状态识别算法，通过数据的实时采集，处理器利用运动识别算法，根据运动数据的特征和实时心率数据对健身动作进行识别，并很清楚的识别出快跑和慢跑，可以提高健身人群的健身效率，更好、更方便的指导健身人群的训练。
附图说明
图1是本申请实施例的健身动作识别方法的流程图；
图2为波比跳动作特征示意图；
图3为引体向上动作特征示意图
图4为深蹲动作特征示意图；
图5为仰卧起坐动作特征示意图；
图6为高抬腿动作特征示意图；
图7为开合跳动作特征示意图；
图8为硬拉动作特征示意图；
图9为跑步动作特征示意图；
图10为本申请实施例的健身动作识别***的硬件***框架图；
图11是本申请实施例的健身动作识别***的结构示意图；
图12是本申请实施例提供的健身动作识别方法的硬件设备结构示意图。
具体实施方式
为了使本申请的目的、技术方案及优点更加清楚明白，以下结合附图及实 施例，对本申请进行进一步详细说明。应当理解，此处所描述的具体实施例仅用以解释本申请，并不用于限定本申请。
请参阅图1，是本申请实施例的健身动作识别方法的流程图。本申请实施例的健身动作识别方法包括以下步骤：
步骤100：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数据(加速度、角速度、磁强度等)和心率数据；
步骤100中，运动数据采集由STM32和MPU9250完成，STM32和MPU9250两者通过IIC总线连接，MCU通过MPU9250相应的寄存器进行设置，包括采样率、传感器量程等寄存器，本申请实施例中，默认加速度量程为±8g，陀螺仪为±1000dbps，磁力计工作在单次测量模式，具体可根据实际操作进行设定。各传感器采样一次可输出6个字节的数据，每个传感器的三轴的输出各占2个字节，高位在前。心率数据采集由STM32和心率传感器来完成，心率传感器通过IIC总线与STM32连接，并进行其寄存器的配置。
步骤110：对采集的心率数据进行滤波处理，去除运动伪迹，得到实时心率值；
步骤110中，心率值计算方法为：
最大运动心率＝(220-现在年龄)*0.8；
最小运动心率＝(220-现在年龄)*0.6；
静息心率正常成年人一般为60-100次/分，当人体处于静息状态时，根据心率传感器数据，每10秒记录一次(h i)，连续记录5组，求平均值，然后乘以6，即得到每分钟静息心率(heart)：
Figure PCTCN2019130588-appb-000001
步骤120：对采集的运动数据进行数据校准和滤波处理，得到三轴加速度、 三轴角速度和三轴磁力计数据；
步骤130：将三轴加速度、三轴角速度和三轴磁力计数据进行融合，得到合加速度、合角速度以及姿态解算所需要的四元数；
步骤130中，数据融合的目的是为了得到姿态解算所需要的四元数，四元数计算量小，无奇点且可以满足飞行器运动过程中姿态的实时解算。对于一个确定的向量，当用不同的坐标系表示时，它们所表示的大小和方向一定是相同的，但是由于两个坐标系的旋转矩阵存在误差，当一个向量经过有误差存在的旋转矩阵后，在另一个坐标系中和理论值会存在偏差，***可以通过这个偏差来修正这个旋转矩阵，旋转矩阵的元素就是四元数，修正后的四元数就能转换为误差较小的姿态角。
三轴加速度值Accx、Accy、Accz、合加速度Accsum：
Figure PCTCN2019130588-appb-000002
三轴角速度Gyrx、Gyry、Gyrz、合角速度Gyrsum：
Figure PCTCN2019130588-appb-000003
步骤140：对四元数进行转换，分别得到姿态角Pitch(俯仰角)、Roll(横滚角)和Yaw(航向角)数据。
步骤150：通过合加速度、合角速度和横滚角(Roll)的特征以及实时心率值对健身动作进行识别；
步骤150中，每一个健身动作对应的合加速度、合角速度和横滚角(Roll)的特征以及心率值都有所不同，以下分别以波比跳、引体向上、深蹲、仰卧起坐、高抬腿、开合跳、硬拉、跑步(快跑和慢跑)动作进行具体说明。具体如图2至图9所示，分别为波比跳、引体向上、深蹲、仰卧起坐、高抬腿、开合跳、硬拉、跑步(快跑和慢跑)的动作特征示意图。如图2所示，每一个波比 跳动作的完成，合加速度会出现四个波峰，Roll角会出现两个波谷；如图3所示，是在实验中采集的三个引体向上，可以看出合加速度与合角速度都有三个波峰。如图4所示，每一个深蹲动作的完成，合角速度都会出现两个波峰，Roll角会同步出现一个波峰。如图5所示，每一个仰卧起坐动作的完成，Roll角都会出现一个波峰，与此同时合角速度会出现两个连续的波峰；如图6所示，每一次高抬腿动作的完成，合加速度和横滚角都会出现一次时间间隔短的波峰；如图7所示，每一次开合跳的完成，合加速度都会出现一个波峰；如图8所示，每一个硬拉动作的完成，Roll角都会出现波谷，与此同时合角速度会出现两个波峰；如图9所示，跑步时合加速度会周期性的出现波峰；在实验中分别采集一组快跑和慢跑的心率，当快跑时心率达到125次/分，慢跑时心率99次/分，因此结合实时心率值即可清楚的识别出快跑和慢跑。
步骤160：根据健身动作识别结果进行相应的计时/计数，并根据设定的时间/次数阈值进行提醒操作。
步骤160中，以波比跳、引体向上、深蹲、仰卧起坐、高抬腿、开合跳、硬拉、跑步动作为例，在健身动作识别结果为波比跳、开合跳、高抬腿或跑步时进行计时，并在计时到达设定的计时阈值(本申请实施例中，计时阈值设定为一分钟，具体可根据实际操作进行设定)时提醒一次；在健身动作识别结果为硬拉、引体向上、深蹲或仰卧起坐时进行计数，并在计数到达设定的计数阈值(本申请实施例中，计数阈值设定为10次，具体可根据实际操作进行设定)时提醒一次。
请参阅图10，为本申请实施例的健身动作识别***的硬件***框架图。硬件***包括惯性传感器模块、心率传感器模块、USB转换模块、固件下载接口、USB供电接口和主控模块。其中，主控模块采用STM32F407ZGT6芯片，其主 频高达168MHZ，1MB的FLASH、192KB的SRAM为运行可靠稳定的无线传感器网络程序以及实现数据高速实时存储提供了快速的运算和处理能力，LQFP144超小封装，实现整个传感器节点的微型化。高达14个定时器，3个IIC接口，3个SPI接口，6个USART接口，3个ADC，2个DAC，112个通用IO口等为连接***设备提供了极其丰富的数据通信接口，主控模块内置JTAG接口，通过固件下载接口即可下载和调试程序。
USB转换模块的芯片为CP2102，其与主控模块的通信协议为USART，具有集成度高的特点，可内置USB2.0全速功能控制器、USB收发器、晶体振荡器、EEPROM及异步串行数据总线(UART)，支持调制解调器全功能信号，无需任何外部的USB器件，可以完成传感器网络节点的USART接口的RS232协议和USB2.0协议的电平转换和通信控制的工作。
惯性传感器模块作为***的数据来源，IMU(Inertial Measurement Unit)需要具有高可靠性、高稳定性以及抗干扰能力。MPU9250集成了3轴加速度、3轴陀螺仪和数字运动处理器(DMP)，可直接通过SPI或I2C输出9轴的全部数据。九轴数据的量程可编程。芯片采用QFN封装，有利于减小整个***的体积，多量程可选能满足***对人体各种动作数据的采集要求，DMP为其提供了多种数据融合的方式；低功耗模式能够在静态时降低***功耗，满足***对低功耗的要求。
请参阅图11，是本申请实施例的健身动作识别***的结构示意图。本申请实施例的健身动作识别***包括惯性传感器模块、心率传感器模块、运动识别算法模块、健身动作识别模块和健身提醒模块。
惯性传感器模块：用于通过九轴惯性传感器采集人体运动时的运动数据(加速度、角速度、磁强度等)；其中，运动数据采集由STM32和MPU9250 完成，STM32和MPU9250两者通过IIC总线连接，MCU通过MPU9250相应的寄存器进行设置，包括采样率、传感器量程等寄存器，本申请实施例中，默认加速度量程为±8g，陀螺仪为±1000dbps，磁力计工作在单次测量模式，具体可根据实际操作进行设定。各传感器采样一次可输出6个字节的数据，每个传感器的三轴的输出各占2个字节，高位在前。
心率传感器模块：用于通过心率传感器采集人体运动时的心率数据；其中，心率数据采集由STM32和心率传感器来完成，心率传感器通过IIC总线与STM32连接，并进行其寄存器的配置。
运动识别算法模块：用于利用运动识别算法，根据运动数据和心率数据计算得到九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；具体的，运动识别算法模块包括：
心率数据处理单元：用于对采集的心率数据进行滤波处理，去除运动伪迹，得到实时心率值；其中，心率值计算方法为：
最大运动心率＝(220-现在年龄)*0.8；
最小运动心率＝(220-现在年龄)*0.6；
静息心率正常成年人一般为60-100次/分，当人体处于静息状态时，根据心率传感器数据，每10秒记录一次(h i)，连续记录5组，求平均值，然后乘以6，即得到每分钟静息心率(heart)：
Figure PCTCN2019130588-appb-000004
运动数据处理单元：用于对采集的运动数据进行数据校准和滤波处理，得到三轴加速度、三轴角速度和三轴磁力计数据；
数据融合单元：用于将三轴加速度、三轴角速度和三轴磁力计数据进行融合，得到合加速度、合角速度以及姿态解算所需要的四元数；其中，数据融合 的目的是为了得到姿态解算所需要的四元数，四元数计算量小，无奇点且可以满足飞行器运动过程中姿态的实时解算。对于一个确定的向量，当用不同的坐标系表示时，它们所表示的大小和方向一定是相同的，但是由于两个坐标系的旋转矩阵存在误差，当一个向量经过有误差存在的旋转矩阵后，在另一个坐标系中和理论值会存在偏差，***可以通过这个偏差来修正这个旋转矩阵，旋转矩阵的元素就是四元数，修正后的四元数就能转换为误差较小的姿态角。
三轴加速度值Accx、Accy、Accz、合加速度Accsum：
Figure PCTCN2019130588-appb-000005
三轴角速度Gyrx、Gyry、Gyrz、合角速度Gyrsum：
Figure PCTCN2019130588-appb-000006
数据转换单元：用于对四元数进行转换，分别得到姿态角Pitch(俯仰角)、Roll(横滚角)和Yaw(航向角)数据。
健身动作识别模块：用于通过合加速度、合角速度和横滚角(Roll)的特征以及实时心率值对健身动作进行识别；其中，每一个健身动作对应的合加速度、合角速度和横滚角(Roll)的特征以及心率值都有所不同，以下分别以波比跳、引体向上、深蹲、仰卧起坐、高抬腿、开合跳、硬拉、跑步(快跑和慢跑)动作进行具体说明。具体如图2至图9所示，分别为波比跳、引体向上、深蹲、仰卧起坐、高抬腿、开合跳、硬拉和跑步的动作特征示意图。如图2所示，每一个波比跳动作的完成，合加速度会出现四个波峰，Roll角会出现两个波谷；如图3所示，是在实验中采集的三个引体向上，可以看出合角速度有三个波峰。如图4所示，每一个深蹲动作的完成，合角速度都会出现一个波峰，Roll角也会同步出现一个波峰。如图5所示，每一个仰卧起坐动作的完成，Roll角都会出现一个波峰，与此同时合角速度会出现两个连续的波峰；如图6所示，每一 次高抬腿动作的完成，合加速度都会出现一次时间间隔短的波峰；如图7所示，每一次开合跳的完成，合加速度都会出现一个波峰；如图8所示，每一个硬拉动作的完成，Roll角都会出现波谷，与此同时合角速度会出现波峰；如图9所示，跑步时合加速度会周期性的出现波峰；在实验中分别采集一组快跑和慢跑的心率，当快跑时心率达到125次/分，慢跑时心率99次/分，因此结合实时心率值即可清楚的识别出快跑和慢跑。
健身提醒模块：用于根据健身动作识别结果进行相应的计时/计数，并根据设定的时间/次数阈值进行提醒操作。其中，以波比跳、引体向上、深蹲、仰卧起坐、高抬腿、开合跳、硬拉、跑步动作为例，在健身动作识别结果为波比跳、开合跳、高抬腿或跑步时进行计时，并在计时到达设定的计时阈值(本申请实施例中，计时阈值设定为一分钟，具体可根据实际操作进行设定)时提醒一次；在健身动作识别结果为硬拉、引体向上、深蹲或仰卧起坐时进行计数，并在计数到达设定的计数阈值(本申请实施例中，计数阈值设定为10次，具体可根据实际操作进行设定)时提醒一次。
图12是本申请实施例提供的健身动作识别方法的硬件设备结构示意图。如图12所示，该设备包括一个或多个处理器以及存储器。以一个处理器为例，该设备还可以包括：输入***和输出***。
处理器、存储器、输入***和输出***可以通过总线或者其他方式连接，图12中以通过总线连接为例。
存储器作为一种非暂态计算机可读存储介质，可用于存储非暂态软件程序、非暂态计算机可执行程序以及模块。处理器通过运行存储在存储器中的非暂态软件程序、指令以及模块，从而执行电子设备的各种功能应用以及数据处理，即实现上述方法实施例的处理方法。
存储器可以包括存储程序区和存储数据区，其中，存储程序区可存储操作***、至少一个功能所需要的应用程序；存储数据区可存储数据等。此外，存储器可以包括高速随机存取存储器，还可以包括非暂态存储器，例如至少一个磁盘存储器件、闪存器件、或其他非暂态固态存储器件。在一些实施例中，存储器可选包括相对于处理器远程设置的存储器，这些远程存储器可以通过网络连接至处理***。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
输入***可接收输入的数字或字符信息，以及产生信号输入。输出***可包括显示屏等显示设备。
所述一个或者多个模块存储在所述存储器中，当被所述一个或者多个处理器执行时，执行上述任一方法实施例的以下操作：
步骤a：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数据和心率数据；
步骤b：利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
步骤c：根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。
上述产品可执行本申请实施例所提供的方法，具备执行方法相应的功能模块和有益效果。未在本实施例中详尽描述的技术细节，可参见本申请实施例提供的方法。
本申请实施例提供了一种非暂态(非易失性)计算机存储介质，所述计算机存储介质存储有计算机可执行指令，该计算机可执行指令可执行以下操作：
步骤a：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数 据和心率数据；
步骤b：利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
步骤c：根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。
本申请实施例提供了一种计算机程序产品，所述计算机程序产品包括存储在非暂态计算机可读存储介质上的计算机程序，所述计算机程序包括程序指令，当所述程序指令被计算机执行时，使所述计算机执行以下操作：
步骤a：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数据和心率数据；
步骤b：利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
步骤c：根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。
本申请实施例的健身动作识别方法、***及电子设备通过在人的身上佩戴九轴惯性传感器和心率传感器等设备收集运动数据和心率数据，通过运动数据和心率数据设计运动状态识别算法，通过数据的实时采集，处理器利用运动识别算法，根据运动数据的特征和实时心率数据对健身动作进行识别，并很清楚的识别出快跑和慢跑，可以提高健身人群的健身效率，更好、更方便的指导健身人群的训练。
对所公开的实施例的上述说明，使本领域专业技术人员能够实现或使用本申请。对这些实施例的多种修改对本领域的专业技术人员来说将是显而易见的，本申请中所定义的一般原理可以在不脱离本申请的精神或范围的情况下，在 其它实施例中实现。因此，本申请将不会被限制于本申请所示的这些实施例，而是要符合与本申请所公开的原理和新颖特点相一致的最宽的范围。

Claims (11)

  1. 一种健身动作识别方法，其特征在于，包括以下步骤：
    步骤a：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数据和心率数据；
    步骤b：利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
    步骤c：根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。
  2. 根据权利要求1所述的健身动作识别方法，其特征在于，在所述步骤a中，所述利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值具体包括：对采集的心率数据进行滤波处理，去除运动伪迹，得到实时心率值，所述实时心率值包括最大运动心率、最小运动心率和静息心率。
  3. 根据权利要求2所述的健身动作识别方法，其特征在于，在所述步骤a中，所述利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值还包括：对采集的运动数据进行数据校准和滤波处理，得到三轴加速度、三轴角速度和三轴磁力计数据；将三轴加速度、三轴角速度和三轴磁力计数据进行融合，得到合加速度、合角速度以及姿态解算所需要的四元数。
  4. 根据权利要求3所述的健身动作识别方法，其特征在于，在所述步骤a中，所述利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值还包括：将所述三轴加 速度、三轴角速度和三轴磁力计数据进行融合，得到合加速度、合角速度以及姿态解算所需要的四元数；并对所述四元数进行转换，分别得到姿态角、横滚角和航向角数据。
  5. 根据权利要求1至4任一项所述的健身动作识别方法，其特征在于，所述步骤c后还包括：根据所述健身动作识别结果对健身动作进行计时或计数，并根据设定的时间阈值或次数阈值进行提醒操作。
  6. 一种健身动作识别***，其特征在于，包括：
    惯性传感器模块：用于通过九轴惯性传感器采集人体运动时的运动数据；
    心率传感器模块：用于通过心率传感器采集人体运动时的心率数据；
    运动识别算法模块：用于利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
    健身动作识别模块：用于根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。
  7. 根据权利要求6所述的健身动作识别***，其特征在于，所述运动识别算法模块还包括：
    心率数据处理单元：用于对采集的心率数据进行滤波处理，去除运动伪迹，得到实时心率值，所述实时心率值包括最大运动心率、最小运动心率和静息心率。
  8. 根据权利要求7所述的健身动作识别***，其特征在于，所述运动识别算法模块还包括：
    运动数据处理单元：用于对采集的运动数据进行数据校准和滤波处理，得到三轴加速度、三轴角速度和三轴磁力计数据；
    数据融合单元：用于将三轴加速度、三轴角速度和三轴磁力计数据进行融合，得到合加速度、合角速度以及姿态解算所需要的四元数。
  9. 根据权利要求8所述的健身动作识别***，其特征在于，所述运动识别算法模块还包括：
    数据转换单元：用于对所述四元数进行转换，分别得到姿态角、横滚角和航向角数据。
  10. 根据权利要求6至9任一项所述的健身动作识别***，其特征在于，还包括：
    健身提醒模块：用于根据所述健身动作识别结果对健身动作进行计时或计数，并根据设定的时间阈值或次数阈值进行提醒操作。
  11. 一种电子设备，包括：
    至少一个处理器；以及
    与所述至少一个处理器通信连接的存储器；其中，
    所述存储器存储有可被所述一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行上述1至5任一项所述的健身动作识别方法的以下操作：
    步骤a：通过九轴惯性传感器和心率传感器分别采集人体运动时的运动数据和心率数据；
    步骤b：利用运动识别算法，根据所述运动数据和心率数据计算得到所述九轴惯性传感器的合加速度、合角速度、横滚角以及实时心率值；
    步骤c：根据所述九轴惯性传感器的合加速度、合角速度和横滚角的特征以及实时心率值对健身动作进行识别。
PCT/CN2019/130588 2019-04-10 2019-12-31 一种健身动作识别方法、***及电子设备 WO2020207071A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/481,323 US20220001262A1 (en) 2019-04-10 2021-09-22 Fitness motion recognition method and system, and electronic device

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910285229.7A CN111803903A (zh) 2019-04-10 2019-04-10 一种健身动作识别方法、***及电子设备
CN201910285229.7 2019-04-10

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/481,323 Continuation US20220001262A1 (en) 2019-04-10 2021-09-22 Fitness motion recognition method and system, and electronic device

Publications (1)

Publication Number Publication Date
WO2020207071A1 true WO2020207071A1 (zh) 2020-10-15

Family

ID=72751515

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/130588 WO2020207071A1 (zh) 2019-04-10 2019-12-31 一种健身动作识别方法、***及电子设备

Country Status (3)

Country Link
US (1) US20220001262A1 (zh)
CN (1) CN111803903A (zh)
WO (1) WO2020207071A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113181619A (zh) 2021-04-09 2021-07-30 青岛小鸟看看科技有限公司 一种运动训练方法、装置和***
CN114011045A (zh) * 2021-11-04 2022-02-08 深圳市云蜂智能有限公司 一种基于可穿戴设备的健身动作计数方法及可穿戴设备
CN114028784B (zh) * 2021-11-25 2023-01-31 深圳先进技术研究院 应用于链球训练的穿戴式运动生物信息监测***及方法
CN114425149A (zh) * 2021-12-13 2022-05-03 中国船舶重工集团公司第七一六研究所 一种仰卧起坐运动计数装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090148000A1 (en) * 2003-12-11 2009-06-11 Nels Howard Madsen System and Method for Motion Capture
CN103417201A (zh) * 2013-08-06 2013-12-04 中国科学院深圳先进技术研究院 一种采集人体姿态的体育运动辅助训练***及其实现方法
CN104571506A (zh) * 2014-12-25 2015-04-29 西安电子科技大学 一种基于动作识别的智能手表及动作识别方法
CN106108909A (zh) * 2016-06-14 2016-11-16 夏烬楚 一种人体姿态检测穿戴设备、***及控制方法
CN106563260A (zh) * 2016-10-28 2017-04-19 深圳职业技术学院 基于姿态传感器的乒乓球智能运动***及其计算方法

Family Cites Families (46)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9900669B2 (en) * 2004-11-02 2018-02-20 Pierre Touma Wireless motion sensor system and method
CN101953682A (zh) * 2010-10-15 2011-01-26 张辉 基于袖带装置的心率检测方法
US9524424B2 (en) * 2011-09-01 2016-12-20 Care Innovations, Llc Calculation of minimum ground clearance using body worn sensors
DE102011121259B3 (de) * 2011-12-15 2013-05-16 Fabian Walke Verfahren und Vorrichtung zur mobilen Trainingsdatenerfassung und Analyse von Krafttraining
CN102814037A (zh) * 2012-08-28 2012-12-12 洪德伟 一种健身运动提示方法及相关装置
US20160051167A1 (en) * 2012-10-10 2016-02-25 Invensense, Inc. System and method for activity classification
WO2014100045A1 (en) * 2012-12-17 2014-06-26 Qi2 ELEMENTS II, LLC Foot-mounted sensor systems for tracking body movement
US20140278219A1 (en) * 2013-03-15 2014-09-18 Focus Ventures, Inc. System and Method For Monitoring Movements of a User
US9213889B2 (en) * 2013-03-28 2015-12-15 The Regents Of The University Of Michigan Athlete speed prediction method using data from attached inertial measurement unit
US10563981B1 (en) * 2013-08-22 2020-02-18 Moov Inc. Automated motion data processing
WO2015048884A1 (en) * 2013-10-03 2015-04-09 Push Design Solutions, Inc. Systems and methods for monitoring lifting exercises
US9681827B2 (en) * 2013-10-09 2017-06-20 LEDO Networks, Inc. Systems, methods, applications for smart sensing, motion activity monitoring, and motion activity pattern recognition
GB2519987B (en) * 2013-11-04 2021-03-03 Imperial College Innovations Ltd Biomechanical activity monitoring
US10417900B2 (en) * 2013-12-26 2019-09-17 Intel Corporation Techniques for detecting sensor inputs on a wearable wireless device
US9239571B2 (en) * 2014-04-25 2016-01-19 Thomas Patton Workout cycle employed in a time measurement portable device
CN106659928B (zh) * 2014-04-29 2020-03-17 特里通韦尔公司 无线度量计算和反馈装置、***和方法
KR101604024B1 (ko) * 2014-06-09 2016-03-28 주식회사 앱스톤 운동량 표시 장치 및 방법
US10652696B2 (en) * 2014-07-30 2020-05-12 Trusted Positioning, Inc. Method and apparatus for categorizing device use case for on foot motion using motion sensor data
US10274318B1 (en) * 2014-09-30 2019-04-30 Amazon Technologies, Inc. Nine-axis quaternion sensor fusion using modified kalman filter
US10065074B1 (en) * 2014-12-12 2018-09-04 Enflux, Inc. Training systems with wearable sensors for providing users with feedback
CN104461013B (zh) * 2014-12-25 2017-09-22 中国科学院合肥物质科学研究院 一种基于惯性传感单元的人体动作重构与分析***及方法
CN105832315A (zh) * 2015-01-16 2016-08-10 中国科学院上海高等研究院 不受个体运动状态与环境位置影响的远程监测***
CN105068654B (zh) * 2015-08-14 2018-05-25 济南中景电子科技有限公司 基于can总线和惯性传感器的动作捕获***及方法
US10172541B2 (en) * 2015-09-25 2019-01-08 J-Mex Inc. Motion recognition device and method
US10314520B2 (en) * 2015-10-02 2019-06-11 Seismic Holdings, Inc. System and method for characterizing biomechanical activity
CN106618542A (zh) * 2015-10-28 2017-05-10 中国科学院上海高等研究院 一种去噪心率信号检测装置及方法
US20170259120A1 (en) * 2016-03-08 2017-09-14 Your Trainer Inc. Programming environment for adaptive workout video composition
US10716495B1 (en) * 2016-03-11 2020-07-21 Fortify Technologies, LLC Accelerometer-based gait analysis
JP2017176207A (ja) * 2016-03-28 2017-10-05 セイコーエプソン株式会社 パフォーマンス情報通知装置、パフォーマンス情報通知方法、およびパフォーマンス情報通知プログラム
CN105929940B (zh) * 2016-04-13 2019-02-26 哈尔滨工业大学深圳研究生院 基于细分特征值法的快速三维动态手势识别方法及***
US20170337033A1 (en) * 2016-05-19 2017-11-23 Fitbit, Inc. Music selection based on exercise detection
CN107478223A (zh) * 2016-06-08 2017-12-15 南京理工大学 一种基于四元数和卡尔曼滤波的人体姿态解算方法
WO2017217567A1 (ko) * 2016-06-15 2017-12-21 (주)그린콤 피트니스 모니터링 시스템
US10918907B2 (en) * 2016-08-14 2021-02-16 Fitbit, Inc. Automatic detection and quantification of swimming
WO2018094011A1 (en) * 2016-11-16 2018-05-24 Lumo Bodytech, Inc. System and method for personalized exercise training and coaching
KR101881028B1 (ko) * 2016-11-21 2018-08-17 홍대건 운동 유형 인식 장치 및 방법
CN106725381B (zh) * 2016-12-19 2019-12-24 华南农业大学 一种智能健身运动手环
CN107016342A (zh) * 2017-03-06 2017-08-04 武汉拓扑图智能科技有限公司 一种动作识别方法及***
CN107273857B (zh) * 2017-06-19 2021-03-02 深圳市酷浪云计算有限公司 运动动作的识别方法及装置、电子设备
US11690535B2 (en) * 2017-07-27 2023-07-04 Universiteit Gent Mobile system allowing adaptation of the runner's cadence
CN108225370B (zh) * 2017-12-15 2024-01-30 路军 一种运动姿态传感器的数据融合与解算方法
EP3797299A4 (en) * 2018-04-26 2022-01-26 Omnibus 157 Pty Limited SYSTEMS AND METHODS FOR FORMULATING A SWIMMER'S MOVEMENT PERFORMANCE MEASURE
CN108703760A (zh) * 2018-06-15 2018-10-26 安徽中科智链信息科技有限公司 基于九轴传感器的人体运动姿态识别***及方法
CN109260673A (zh) * 2018-11-27 2019-01-25 北京羽扇智信息科技有限公司 一种运动计数方法、装置、设备及存储介质
US20200261023A1 (en) * 2019-02-14 2020-08-20 Athletai Co. Ascertaining, Reporting, and Influencing Physical Attributes And Performance Factors of Athletes
US11253747B2 (en) * 2019-03-07 2022-02-22 Bose Corporation Automated activity detection and tracking

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090148000A1 (en) * 2003-12-11 2009-06-11 Nels Howard Madsen System and Method for Motion Capture
CN103417201A (zh) * 2013-08-06 2013-12-04 中国科学院深圳先进技术研究院 一种采集人体姿态的体育运动辅助训练***及其实现方法
CN104571506A (zh) * 2014-12-25 2015-04-29 西安电子科技大学 一种基于动作识别的智能手表及动作识别方法
CN106108909A (zh) * 2016-06-14 2016-11-16 夏烬楚 一种人体姿态检测穿戴设备、***及控制方法
CN106563260A (zh) * 2016-10-28 2017-04-19 深圳职业技术学院 基于姿态传感器的乒乓球智能运动***及其计算方法

Also Published As

Publication number Publication date
CN111803903A (zh) 2020-10-23
US20220001262A1 (en) 2022-01-06

Similar Documents

Publication Publication Date Title
WO2020207071A1 (zh) 一种健身动作识别方法、***及电子设备
EP2802255B1 (en) Activity classification in a multi-axis activity monitor device
US20190328243A1 (en) Methods and Systems for Determining Abnormal Cardiac Activity
US11540761B2 (en) Systems and methods for facilitating exercise monitoring with real-time heart rate monitoring and motion analysis
CN106725381B (zh) 一种智能健身运动手环
CN111475024B (zh) 一种人体动作捕捉***及方法
CN107754225A (zh) 一种智能健身教练***
CN108366742B (zh) 一种生物信号采集方法、装置、电子设备及***
Yang et al. Real-time human activity classification by accelerometer embedded wearable devices
CN212729816U (zh) 一种人体动作捕捉***
CN109758154A (zh) 一种运动状态确定方法、装置、设备及存储介质
CN205375758U (zh) 一种老年人主动急救***
CN210078765U (zh) 基于可穿戴传感器的动作捕捉识别与评估装置
CN104808778B (zh) 判断头戴式智能设备操作有效性的装置和方法
Prudêncio et al. Physical activity recognition from smartphone embedded sensors
WO2017172105A1 (en) Opportunistic measurements and processing of user's context
CN110624217A (zh) 一种基于多传感器融合的复健手套及其实现方法
Qin et al. Real-time monitoring system of exercise status based on internet of health things using safety architecture model
Okamoto et al. Design of wireless waist-mounted vital sensor node for athletes—Performance evaluation of microcontrollers suitable for signal processing of ECG signal at waist part
CN207168915U (zh) 一种新型多功能跳绳
CN203122370U (zh) 一种人体健康指数综合评测装置
Zhang Smartphone based activity recognition system
Tsukahara et al. A 15-μA metabolic equivalents monitoring system using adaptive acceleration sampling and normally off computing
Lovell A system for real-time gesture recognition and classification of coordinated motion
CN210843020U (zh) 一种基于nfc技术的人体心率检测装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19924412

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19924412

Country of ref document: EP

Kind code of ref document: A1