Enhanced gait recognition using HMM and Vh techniques

Vision-based human identification at a distance in surveillance has recently gained more attentions. Gait has the advantages of being non-invasive and difficult to conceal, and is also the only perceivable biometric at a distance. A novel method for enhancing the gait recognition rate is proposed by combining HMM method and the visual hull method to identify and authenticate the gait of a person. By using polyhedral rendering for visual hull process, the interaction error is reduced due to Kalman filtering. The values between centroid and other identified regions are obtained. The obtained values are in 2D and the HMM technique is used to identify the distance and these values are stored in the database. The advantage of this method is that the value of the centroid is accurate. Hence, authentication is more accurate since most of the distance identification is made with reference to the centroid.