Gait Recognition from Front and Back View Sequences Captured Using Kinect Pratik Chattopadhyay 1 , Shamik Sural 1 , and Jayanta Mukherjee 2 1 School of Information Technology, IIT Kharagpur, India 2 Dept. of Computer Science & Engineering, IIT Kharagpur, India {pratikc,shamik@sit.iitkgp.ernet.in, jay@cse.iitkgp.ernet.in Abstract. In this paper, we propose a key pose based gait recognition approach using skeleton joint information derived from the depth data of Kinect. We consider situations where such depth cameras are mounted on top of entry and exit points, respectively capturing back and front views of subjects who enter a zone under surveillance. Three dimensional geometric transformations are used to map the skeleton images captured from the back view to an equivalent front view. A gait cycle is divided into a number of key poses and the trajectory followed by each skeleton joint within a key pose is used to derive the gait features for that par- ticular pose. For recognizing a subject, available key poses are compared with the corresponding key poses of the training subjects. The proposed method has higher accuracy than other competing approaches. Keywords: Gait Recognition, Kinect, Geometric Transformation, Skeleton Joint, Key Pose. 1 Introduction and Background Gait recognition is the process of identifying an individual from his style of walking. Suitability of gait as a biometric feature in security applications is established by the fact that unlike other biometric methods like face detection, iris scan and finger print detection, close interaction with a subject is not required in gait recognition. Gait can be identified from a distance and even with low resolution video sequences. Most of the existing work on gait recognition considers capturing of video sequences in the fronto-parallel (side) view, since maximum gait information during walking is contained in this view [1], [2]. It has, however, been pointed out that the assumption on the availability of gait features from the fronto- parallel view is too restrictive especially in situations where gait is presumed to be the biometric of choice, i.e., in places like airports and railway stations [3]. In such places, people are often directed to enter into secured zones one at a time like walking through a metal detector gate. Since it is easier to control the flow of persons at the entrance/exit of a security zone, surveillance cameras mounted above the entry/exit point can record a better view of a subject entering/leaving P. Maji et al. (Eds.): PReMI 2013, LNCS 8251, pp. 196–203, 2013. c  Springer-Verlag Berlin Heidelberg 2013