Gait Feature Extraction Using Dyadic Wavelet Transform and Structural Similarity for Gait Recognition

摘要

51.1 Introduction Gait recognition is a biometric authentication method for identifying an individual that is focused on his/her manner of walking. In contrast to biometric authentication methods that use a fingerprint or an iris, where the subject must be close to the sensor for recognition processing, gait recognition methods can recognize a human being at a long distance from the monitoring camera video. Gait recognition methods authenticate the individual person based on the gait features of the object person extracted from the image sequence obtained from the monitoring camera video.