A Method of DTW Based Gait Recognition and Gait Data from Kinect

Gait recognition as a field of biometrics more and more attention. Many people are interested in working with Kinect to study gait recognition. In this paper, we propose a Dynamic Time Warping (DTW) algorithm to classify the features by using the inherent physiological characteristics of human and Kinect's skeletal tracking function, extracting the distance between skeletal and joint points, the height of body mass center, height and centroid to the main skeletal joint points . The experimental results show that this method can obtain more than 90% recognition rate in the UPCV database and has good robustness.