Design And Implementation Of Indoor Environment Human Tracking And Gait Recognition System Based On Millimeter Wave Radar

With the development of 5G technology and Io T,smart home devices have begun to enter more and more homes.Smart home devices usually acquire the ability to perceive the home environment or family members through visual means.But the visual solution has limitations for the home environment.With the development of FMCW technology,it has become more and more feasible to use millimeter wave radar for home environment perception.A system for human tracking and gait recognition in indoor environments is constructed by collecting data from millimeter-wave radar.Based on the principle of millimeter wave radar,this paper designs a millimeter wave radar point cloud acquisition module,and based on the analysis of the characteristics of millimeter wave point clouds,designs a preprocessing module for noise point filtering of the point cloud,and then the processing module will cluster the point cloud through the DBSCAN algorithm,and the human body's motion trajectory is generated by the Hungarian algorithm according to the obtained clusters,and the Kalman filter is used to deal with the situation of occlusion or tracking loss.The gait data of the human body is recognized,and the person to whom the trajectory belongs is determined.In the implementation,data acquisition,data processing,and data display are paralleled by three processes,and the processes communicate through queues.The system uses TI's IWR6843-ISK radar board to collect data and implement subsequent processing through python.The final system can filter out the clutter generated by the radar well,and can accurately track and identify with a small number of people.