Numerical calculus solution of gait recognition algorithm based on compressed sensing

. X-ray stereo imaging analysis technology requires multiple X-ray fluoroscopy of the object of study, and ionizing radiation has potential health hazards. At the same time, pre-implantation of markers and collection of large amounts of data is unrealistic in large-scale studies. Based on this, a three-dimensional numerical calculus model of human knee joint is proposed and introduced into gait recognition to achieve dimension reduction and feature extraction of gait sequence. It firstly quantifies the changes of gait parameters of patients after total knee arthroplasty by adopting three-dimensional gait analysis technique, and constructs a three-dimensional numerical calculus model of the knee joint in this paper. Then the three-dimensional numerical calculus model of knee joint is applied to gait feature extraction. To solve the model, an adaptive selection algorithm based on gait feature dimension is proposed to realize the autonomous selection of gait feature vector dimension. The results show that the low-dimensional observation measurement contains important information for accurately reconstructing the original gait image, which can be used to characterize the image. The training and recognition experiments of gait samples using SVM multi-classifier also verify the validity of the gait features extracted according to the algorithm. The research results show that the proposed method has certain feasibility and there is room for improvement.