The Research And Implement Of A Gait Recognition System

Gait recognition has received growing interest within the computer vision community recently for it is the only one of biometric recognition which can apperceived with long distance. Focusing on gait recognition, this paper describes method of human identification and feature extraction, and the work in the following aspects of the research field is performed.Firstly, this paper describes the method of extracting gait feature, improved background subtraction arithmetic which bring forward using the RGB-sensitive-weights in RGB color space, get the moving human silhouettes in color image sequence by using an algorithm, integrated the algorithms of histogram two threshold segmentation and morphological operators.Secondly, this paper study the method of gait token and gait recognition. We extract the four key frame of gait by the width of silhouettes by analyzing the cyclic gait, then use Fourier descriptors(FD) to process body contour extracting from the key frames, and extract the characteristic data which can provided for recognition. At last, we choose two different classifier(NN classifier and ENN classifier) experiments on NLPR gait-database, then, discusses the result of recognition.Thirdly, we have designed a system of gait recognition, and developed a simply gait recognition system with Microsoft Visual C++6.0 on Windows System.