Research On Human Identification Based On Gait

The identity recognition which is based on gait is a new biology recognition technique in recent years. Human gait recognition is the process of identifying individuals by their walking manners and is attractive in pattern recognition and image processing, and that is only behavioral characteristics of biometrics which can be detected and measured at a distance. Compared with traditional biology recognition, the gait has unexampled advantages such as difficult to disguise and hide, simple collection, identification from for away even in low-resolution and so on.Generally, gait recognition composed of four parts: Preprocessing of gait sequences,Period detection, feature extraction and classification.The preprocessing of gait sequences is the first step of gait recognition. It mainly includes background model, motion detection, binarization and morphologic postprocessing. The mean method is used to restore background and an improved approach of background subtraction is proposed. Finally, Binarization and morphologic algorithm are used to remove image noises. The width information of human is used to detect the gait period is this paper. A consolidation period method is also proposed in this paper in order to convenient for recognition. Feature extraction is the most important thing, on which our study focuses. The Zernike moments can well describe overall feature of image, but fail to describe local feature. According to the advantage of wavelet moments to describe details in local, a new gait recognition method is presented by combining wavelet moments with Zernike moments to describe gait sequence images. Then, a new fusion algorithm is suggested wherein the static and dynamic features are fused to obtain optimal performance. The new fusion algorithm divides decision situations into two categories. The Zernike moment is used to describe the static features of gait sequence images, and the three widths of the body contour is used to describe the dynamic features. In the recognition part, Because of the improved BP neural network, we can ensure effectiveness of the classification as well as speed.