Research On The Key Technologies Of Human Posture Analysis And Recognition Based On Inertial Measurement

Human posture analysis and recognition technology has broad application prospects,this paper presents a humanoid attitude analysis and recognition system based on inertial sensors.The system collects body posture information through the nodes distributed in the main joints of the whole body and transmits the posture information to the computer through the wireless transmission.The attitude reconstruction is realized on the computer,and the typical posture is identified,and the related parameters are monitored.The system has features of low cost,lightweight and convenient,adaptable,secondary development and so on.In this paper,we first investigate the background and current situation of posture analysis and recognition technology,research on the basic theories that may be involved,Based on this,a human body attitude analysis and recognition system based on inertial sensors is designed,the key technologies involved in it are studied in depth.Firstly,according to the hardware features of the system and the need of wireless transmission,the real-time transmission technology of multi-node data receiving and transmitting based on ESP8266 wireless module is deeply studied,the hardware and experimental platform of this paper are determined.The inertial sensor calibration technology is studied.In the context of resource-constrained environment and low cost,a gradient-decision-based fast attitude calculation algorithm is designed and implemented.The optimization and cutting are combined with the actual system.For the research of attitude reconstruction technology,a multi-threaded data synchronization and attitude following PC software is designed based on OpenGL.The joint association data of multiple nodes is constrained and reconstructed,and the reconstruction results are displayed through graphic image rendering.Combining related kinematics knowledge to identify postures and monitor key attitude-related indicators.For the entire system,various key technologies are organically integrated,and an attitude resolving analysis and reconstruction recognition system based on a wearable wireless sensor network is designed.Experiments show that the gesture acquired by this system can stably and accurately reconstruct the posture on the computer software,and the recognition of each gesture and parameter achieves high precision.During 30 minutes of constant measurement,the static error is within 1 degree and the average of dynamic error is within 5 degree,it is within the acceptable range,good followability under dynamic conditions.This system overcomes the limitations of the traditional wired or optical methods and can be widely used in the fields of sports training,human-computer interaction control,rehabilitation medical field,games and film and television production.