CN103136912A - Moving posture capture system - Google Patents

Abstract

The motion posture capturing system of the present invention includes a plurality of sensors for measuring motion postures and a host computer for processing posture data, wherein: the inertial posture measurement sensor module is bidirectionally connected with the main control module through a USART communication cable, and the main control module and the upper computer Wireless two-way connection, the power management module is connected to the inertial attitude measurement sensor module and the main control module respectively. The system integrates a three-axis magnetic field sensor and a six-axis acceleration and angular velocity sensor, and uses inertial strapdown attitude calculation, combined with Kalman digital filtering and data fusion algorithms, to achieve accurate three-dimensional measurement and automatically correct the collected data. It can not be affected by environmental factors, and can accurately capture the movements and postures of people, animals and objects in real time. Users can directly use the data output by the system to produce more realistic and high-quality animations. It can be used for dynamic simulation such as character models in animation, film and television, and games, as well as scientific research on the motion state of people, animals, and objects.

Description

Motion gesture capture system

technical field

The invention relates to data collection of motion postures, in particular to a motion posture capture system.

Background technique

3D dynamic/static simulation is becoming more and more important in today's digital age. If the dynamic simulation of animation, film and game role models is made by using physical modeling or manual drawing, it is obviously not suitable for the trend, low efficiency, heavy workload and high cost. . As people's requirements for motion simulation of character models are getting higher and higher, the requirements for the efficiency, convenience of use, precision and cost of motion capture systems are also getting higher and higher, which requires the system to be low-cost, simple and Ease of use, high precision and high efficiency are developed. Motion capture systems are increasingly used in the production of 3D dynamic simulation of animation, film and television and games, and 3D dynamic simulation has become the main force in the 3D manufacturing industry.

Internationally, especially in the United States, where technology is advanced, dynamic simulation of character models has already been developed, and it is reflected in American animated films, science fiction films and games, and has reached a certain level. Through the motion capture system, it is very convenient to realize the actions required by the designer for the character model. You only need to build the character model in the computer, and then install the motion capture system on the actor, so that the actor can do the action required by the designer. Then import the data collected by the motion capture system into the character model through the 3D animation production software, and the character model will make the same motion simulation animation as the actor. Therefore, the three-dimensional dynamic simulation can be realized quickly and conveniently, which reduces the cost and saves time.

There are three main types of motion capture systems at present: 1. Realized by optical principles, but this kind of system is affected by conditions and environmental factors, is extremely sensitive to changes in light, and the operation of the system is also relatively complicated. As mentioned above, it is easily affected by the environment and its use is easily limited; 2. Through the acoustic principle, this method is not very ideal, the anti-interference ability is poor, the system operation is complicated, and it is not easy to realize; 3. Through mechanical sensors, This method is simple and feasible, the algorithm is simple, and it is easier to implement than the previous two, and it is not affected by environmental and geographical factors, and the system operation is also more convenient. There is room for improvement in terms of sex and so on.

Contents of the invention

The purpose of the present invention is to provide a motion posture capture system, which can not be affected by environmental factors, and can accurately capture the motion postures of people, animals and objects in real time, and the computer can directly use the data output by the system to improve the quality of animation , making the effect more realistic. It can serve dynamic simulations such as role models for animation, film and television, and games, as well as scientific research on the motion states of people, animals, and objects.

In order to achieve the above object, the present invention adopts following technical scheme:

The motion posture capture system includes a plurality of sensors for measuring motion posture and a host computer for processing posture data, wherein: the inertial posture measurement sensor module is bidirectionally connected with the main control module through a USART communication cable, and the main control module is wireless with the upper computer Two-way connection, the power management module is connected to the inertial attitude measurement sensor module and the main control module respectively.

Further, the inertial attitude measurement sensor module is connected to the network node data set in the main control module, the node data processing module and the human body attitude data integration module through the USART communication cable.

In order to carry out data acquisition and processing better, in the above-mentioned inertial attitude measurement sensor module, the attitude measurement central processor integrates an internal FLASH parameter reading device, an internal parameter calibration device, a host computer command reading device, a timer, signal acquisition and Filter processing device, data fusion and attitude calculation device, USART communication interface, internal FLASH parameter reading device is connected to internal parameter calibration device, timer, signal acquisition and filter processing device, data fusion and attitude calculation device, USART communication in turn Interface, the host computer command reading device behind the USART communication interface is connected to the internal parameter calibration device, and the three-axis magnetic field sensor, six-axis acceleration and angular velocity sensor are all connected to the signal acquisition and filter processing device of the attitude measurement central processor. Among them, the attitude calculation is best to use inertial strapdown attitude calculation, and the filter processing is best to use Kalman filter algorithm.

In order to transmit data, the above-mentioned three-axis magnetic field sensor, six-axis acceleration and angular velocity sensor are all connected to the A/D conversion interface of the attitude measurement central processing unit.

In the above-mentioned main control module, the node data fusion module is connected after the network node data collection and the node data processing module, the human body posture data integration module is respectively connected with the node data fusion module, the motion state data transmission protocol module, and the motion state data transmission protocol module Two-way connection with the wireless data transmission module 1.

In the present invention, the wireless data transmission module 2 bidirectionally connected with the wireless data transmission module 1 of the main control module in the host computer is bidirectionally connected with the motion state data interpretation module, and the motion state data interpretation module is bidirectionally connected with the host computer management system.

Further, in order to make the motion posture capture system of the present invention work better and obtain a more realistic, smooth and accurate three-dimensional effect, the host computer management system of the present invention sets the main interface of the host computer management system and links to it respectively:

Key position installation position setting subsystem;

Sensor calibration subsystem;

Sensor sensitivity setting subsystem;

Motion posture data recording subsystem;

Virtual character action display subsystem.

In the present invention, the inertial posture measurement sensor module is mainly placed on the main key parts of the human body: left ankle, right ankle, left knee, right knee, left wrist, right wrist, left elbow, right elbow, neck and waist joints of the human body place. The placement direction of the inertial attitude sensor module (mainly each sensor) can be placed differently according to the actual wiring method, but it needs to be set in the network node data collection and node data processing, otherwise there will be some wrong action.

The network node data collection and node data processing module receives the data of each inertial attitude measurement sensor module, sets the installation direction of each sensor, calibrates whether the output of each sensor module is correct, and performs node data processing, marking each key sensor module data for node data fusion.

The power management module is very important. The stability of the power supply is directly related to the correctness of the sensor data. If the output wave of the power supply is too large, it will directly affect the stable operation of the entire motion state capture system. The power management module is powered by a 16V lithium battery. , The capacity of the lithium battery is 4000 mAh, and the switching power supply circuit design is set to ensure the conversion accuracy and real-time performance.

The human body posture data integration module calculates the main movement direction, speed and mode of each joint based on the data of each node according to the movement mode of the human body, and establishes a three-dimensional space XYZ three-dimensional coordinate system, and generates a corresponding movement mode along with the movement of the human body and routes. At the same time, the original data of the inertial attitude measurement sensor module placed on each main key part of the human body is compared with the fused data to determine the correctness of the three axial data.

Node data fusion module: because the sensor data marked with nodes is an inertial device, there are some noises. In the node data fusion, it is necessary to use software filtering to process the collected data. This process uses Kalman filtering to The interference signal and noise in the collected signal are filtered, and the distorted signal is restored to obtain a signal that meets the requirements of human body posture measurement.

The motion state data transmission protocol module encodes the data integrated with the human body motion posture data. The encoding format is in hexadecimal format. Before the motion state data transmission, the data instruction header is added, and the instruction transmission length is marked, and the motion state data is in the middle. , and finally marked with data checksum and end of data transmission. Using this format can ensure the error-free and real-time transmission.

Wireless data transmission module 1 adopts an existing wireless data transmission module with sending and receiving functions. This module uses XBee-PRO 900HP, and a base board of XBee-PRO 900HP is set. In the process, in order to prevent the processor (CPU) from causing data interference to the module, an emitter follower is specially set up to isolate the crosstalk between the processor and the wireless data transmission module.

The wireless data transmission module 2 is the same as the wireless data transmission module 1 .

The motion state data interpretation module interprets the specific meaning of each field according to the transmitted protocol format, and verifies the correctness of the data. If it is correct after verification, it converts the hexadecimal data of each major joint of the human body into ASCII code.

The host computer management system virtual reconstructs the ASCII code data of the main joints of the human body to the characters on the host computer, and displays the state of motion in real time. At the same time, set the sensitivity of the sensor, send instructions for setting the installation position of key parts and calibrating the sensor, and record the motion posture data.

The course of work of the system of the present invention is:

1. Turn on the computer and run the upper computer management system and related programs;

2. Place the inertial attitude measurement sensor module on various key parts of the performer's body, such as

At the left ankle, right ankle, left knee, right knee, left wrist, right wrist, left elbow, right elbow, neck and waist joints of the human body, the main control module is installed in the back center of the actor's body;

3. Use USART two-way communication cables to connect each inertial attitude measurement sensor module with the main control module, and then open the wireless connection between the main control module and the host computer;

4. The performer stands at attention, all turn on the power of the motion posture capture system, the motion posture capture system starts to initialize, the indicator light of the system starts to flash, and the host computer continuously receives and saves the data of the performer’s motion posture. At the same time, the virtual The character motion display subsystem will display the simulated motion.

The positive effect of the present invention is:

1. The present invention adopts ARM processor technology to realize data acquisition, which ensures the accuracy and real-time performance of data acquisition and data fusion;

2. The system integrates a three-axis magnetic field sensor and a six-axis acceleration and angular velocity sensor, and uses inertial strapdown attitude calculation, combined with Kalman digital filtering and data fusion algorithms, to achieve accurate three-dimensional measurement and provide high-precision attitude;

3. The system adopts a data refresh rate of 60Hz, which greatly improves the fluency and fidelity of animation;

4. The system can automatically correct the collected data, has strong error correction ability, ensures the accuracy of long-term running measurement results, and ensures the accuracy and real-time performance of action posture data.

Description of drawings

Fig. 1 is a schematic diagram of the hardware structure of the motion state capture system of the present invention;

Fig. 2 is a schematic structural diagram of an inertial attitude measurement sensor module;

Figure 3 is a block diagram of the upper computer management system.

Detailed ways

The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

As shown in Figure 1, the inertial attitude measurement sensor module is bidirectionally connected to the main control module through the USART communication cable, the main control module is bidirectionally connected to the upper computer wirelessly, and the power management module is connected to the inertial attitude measurement sensor module and the main control module respectively.

The inertial attitude measurement sensor module is connected to the network node data collection in the main control module, the node data processing module and the human body attitude data integration module through the USART communication cable.

See Figure 2. In the inertial attitude measurement sensor module, the attitude measurement central processor integrates an internal FLASH parameter reading device, an internal parameter calibration device, a host computer command reading device, a timer, a signal acquisition and filtering processing device, a data fusion and attitude calculation device, USART communication interface (two-way), the internal FLASH parameter reading device is sequentially connected to the internal parameter calibration device, timer, signal acquisition and filtering processing device, data fusion and attitude calculation device, USART communication interface, and the host computer behind the USART communication interface The command reading device is connected to the internal parameter calibration device, and the three-axis magnetic field sensor, six-axis acceleration and angular velocity sensor are all connected to the signal acquisition and filter processing device of the attitude measurement central processor. The attitude calculation adopts the inertial strapdown attitude calculation, and the filter processing adopts the Kalman filter algorithm.

The three-axis magnetic field sensor, six-axis acceleration and angular velocity sensor are all connected to the A/D conversion interface of the attitude measurement central processing unit.

After the network node data collection and node data processing module, the node data fusion module is connected. The human body posture data integration module is connected with the node data fusion module and the motion state data transmission protocol module in two directions, and the motion state data transmission protocol module is bidirectionally connected with the wireless data transmission module. connect.

The wireless data transmission module 2 in the host computer is bidirectionally connected with the wireless data transmission module 1 of the main control module, and the motion state data interpretation module is bidirectionally connected, and the motion state data interpretation module is bidirectionally connected with the host computer management system.

See Figure 3. The upper computer management system sets the main interface of the upper computer management system and the following five subsystems respectively linked to it:

Key position installation position setting subsystem - set the installation position and direction of the sensor;

Sensor calibration subsystem - including three-axis magnetic field calibration and six-axis acceleration and angular velocity sensors (three-axis accelerometer calibration, three-axis gyroscope calibration);

Sensor sensitivity setting subsystem - including three-axis magnetic field sensitivity setting and six-axis acceleration and angular velocity sensor (three-axis accelerometer sensitivity setting, three-axis gyroscope sensitivity setting);

Movement attitude data recording subsystem - stored in the hard disk;

Virtual character action display subsystem.

In the present invention, the workflow of the inertial attitude measurement sensor module includes the following steps:

1) At the beginning, the attitude measurement CPU is powered on and starts to work, and the watchdog reset system is started;

2) System preparation, internal parameter calibration parameter reading;

3) Read the gyro initialization and magnetic field calibration instructions issued by the host computer, so as to calibrate each sensor;

4) Read the internal FLASH calibration parameters, and use the factory default parameter values for attitude calculation;

5) Turn on the timer interrupt, start the system interrupt inside the ARM processor, perform internal data fusion, attitude calculation, etc.;

6) Signal acquisition and filtering algorithm, according to the output characteristics of each sensor data, use Kalman filtering technology for filtering;

7) Data fusion and attitude calculation, for the six-axis acceleration and angular velocity sensor, the three-axis magnetic field sensor adopts the four-element method to solve the attitude angle;

8) Transmit the attitude angle data to the main control module through the USART communication interface.

The power management module of the present invention is powered by a 16V lithium battery, the lithium battery has a capacity of 4000 mA, and a switching power supply circuit is provided.

The motion state data transmission protocol module encodes the data integrated with the human body motion posture data. The encoding format is in hexadecimal format. Before the motion state data is transmitted, the data instruction header is added, and the instruction transmission length is marked. The middle is the motion state data. Finally, there are data checksums and data transmission end marks. Using this format can ensure the error-free and real-time transmission.

Wireless data transmission module 1 adopts an existing wireless data transmission module with sending and receiving functions. This module uses XBee-PRO 900HP, and a base board of XBee-PRO 900HP is set. In the process, in order to prevent the processor (CPU) from causing data interference to the module, an emitter follower is specially set up to isolate the crosstalk between the processor and the wireless data transmission module.

The wireless data transmission module 2 is the same as the wireless data transmission module 1 .

The motion state data interpretation module interprets the specific meaning of each field according to the transmitted protocol format, and verifies the correctness of the data. After the verification is accurate, it converts the hexadecimal data of each major joint of the human body into ASCII codes.

The host computer management system virtual reconstructs the ASCII code data of the main joints of the human body to the characters on the host computer, and displays the state of motion in real time.

The system uses a refresh rate of 60Hz data.

The research and experiment results of the present invention show that: under the same simulation step length condition, applying the above method has high calculation accuracy in a high dynamic angular motion environment, and the drift error of the pitch angle algorithm caused by the conical motion is also effectively suppressed. The data processing method of the optimal estimation theory and the Kalman filter are used to fuse the velocity information to ensure the accuracy and real-time performance of the measurement accuracy. And the linear Kalman filter is used to provide the minimum variance estimate for the inertial navigation system error, and then the estimated value of these errors is used to correct the attitude control system.

Claims (8)

1. Motion posture capture system, including a plurality of sensors for measuring motion posture and a host computer for processing posture data, it is characterized in that: the inertial posture measurement sensor module is bidirectionally connected with the main control module through the USART communication cable, and the main control module and the upper position The wireless two-way connection between the machine and the power management module is connected to the inertial attitude measurement sensor module and the main control module respectively. 2. The motion capture system according to claim 1, characterized in that: the inertial attitude measurement sensor module is connected to the network node data set in the main control module and the node data processing module and the human body attitude data integration module through the USART communication cable. 3. The motion capture system according to right 1, characterized in that: in the inertial attitude measurement sensor module, the attitude measurement central processing unit integrates internal FLASH parameter reading device, internal parameter calibration device, host computer instruction reading device, timer , signal acquisition and filter processing device, data fusion and attitude calculation device, USART two-way communication interface, the internal FLASH parameter reading device is connected to the internal parameter calibration device, timer, signal acquisition and filter processing device, data fusion and attitude solution in turn After the USART two-way communication interface, the upper computer command reading device is connected to the internal parameter calibration device, and the three-axis magnetic field sensor, six-axis acceleration and angular velocity sensor are all connected to the signal acquisition and processing of the attitude measurement central processor. filter processing device. 4. The motion capture system according to claim 3, characterized in that: the three-axis magnetic field sensor, the six-axis acceleration and angular velocity sensor are all connected to the A/D conversion interface of the attitude measurement central processing unit. 5. The motion capture system according to claim 3, characterized in that: said attitude calculation adopts inertial strapdown attitude calculation, and said filtering process adopts Kalman filter algorithm. 6. The motion capture system according to right 1, characterized in that: in the main control module, the node data fusion module is connected after the network node data collection and the node data processing module, and the human body posture data integration module is respectively connected with the node data fusion module, the motion The state data transmission protocol module is bidirectionally connected, and the motion state data transmission protocol module is bidirectionally connected with the wireless data transmission module 1 . 7. The motion capture system according to right 1, characterized in that: the wireless data transmission module 2 bidirectionally connected with the main control module wireless data transmission module 1 wireless two-way connection motion state data interpretation module in the upper computer, motion state data interpretation module Two-way connection with the host computer management system. 8. The motion capture system according to claim 7, characterized in that: the host computer management system sets the host computer management system main interface and links with it respectively: Key position installation position setting subsystem; Sensor calibration subsystem; Sensor sensitivity setting subsystem; Motion posture data recording subsystem; Virtual character action display subsystem.

Images