A new automatic gait recognition method based on the perceptual curve

Recognizing people by their gait is an emerging biometrics. In this paper we proposed a new automatic gait recognition approach based on the perceptual curve, which can effectively preserve the temporal changes of the walker's silhouette shape consistency with human vision. Firstly, we detect the walking person in each frame of the monocular image sequences. Then we use inner boundary tracking algorithm to extract the walker's binary silhouettes and compute the PSD (Perceptual Shape Descriptor) of the shapes. Next, accumulative PSD of each image sequence forms a perceptual curve as its gait signature. Finally, the STC (Spatio-temporal Correlation) similarity measure and two different simple classification methods (NN and KNN) are used to recognize different subjects. Experimental results on the UCSD database and CMU database demonstrate the feasibility of the proposed approach.