Gait recognition based on the body's natural features

Gait recognition is a new topic in the biometric technology. Human authenticates or identifies individuals by the way they walk. This paper adopts body's natural signatures to recognize individuals. Height, as a static characteristic, and silhouette width, as a dynamic characteristic, contain useful information for identification. Together with them, the gait features are represented. The nearest neighbor classifier is adopted to identify subjects. The proposed algorithm is evaluated on CASIA dataset. The algorithm is simple and fast.