Research On Human Motion Capture Technology Based On MIMU

Human posture analysis and recognition can predict human posture behavior,action classification and abnormal behavior detection,while it is very important for improving human-computer interaction experience,medical health protection,motion analysis,military training normality and other aspects.Along with the development of human posture analysis and recognition technology,many high-risk jobs in life require wearable motion capture devices to avoid accidents to ensure personal safety,which is of great significance to the development of society.Based on the above practical needs,this paper proposes a wearable MIMU human motion capture system based on the extended kalman data fusion algorithm,and carries out the following theoretical analysis and experimental research work:First,based on the analysis of the current domestic and international research status and development trend of wearable human motion capture system,the overall scheme of wearable human motion capture system is designed based on the principle of pose detection and data fusion algorithm,and the number of sensing nodes,wearing position,and the initial human alignment scheme are studied;Then,the data processing platform of the upper computer was designed to receive,parse and extend the kalman filter algorithm data fusion,display,record and transmit the fused data to the TCP module;the motion capture platform was designed to complete the database file extraction,parameter debugging and TCP data control functions using the human model drawn by 3ds Max to realize the human posture following;Finally,four classification algorithms,BP neural network,support vector machine,random forest,and long and short term memory network,are classified and studied for their effectiveness in processing the collected human data set,and the optimal algorithm is selected for experimental study.The experimental analysis of the wearable MIMU human motion capture system’s function,performance,overall average recognition accuracy of posture and other indicators.In this paper,the research of human motion capture technology based on MIMU has the advantages of low power consumption,low cost and high accuracy,which provides experimental basis and technical support for the application of human wearable sensors.