Abstract—In this paper, the adopted methodology based on intersection with fuzzy classification. Consider two human body joint i.e. shoulder and feet. Feet body joint is separated in to two component toe and heel of the left and right leg. Two triangles are formed on joining the shoulder and feet (toe and heel). The intersecting points are computed on CASIA database. Fuzzy classification yield enhanced recognition rate. Index Terms—Gait recognition, intersection, recognition rate, biometric, fuzzy. I. INTRODUCTION Biometrics is derived from Greek word “Bio” means life and “metrics” means measure[1].The term gait was first demonstrate by Johansson in 1970[2]. Some methods such as finger prints and face recognition, already proved to be very efficient in human recognition systems [3]. Face, fingerprint recognition, iris recognition can’t work well from a large distances as well as on degraded images in present scenario security is one of the challenging issues in all field [1],[3]. Human gait is one of the promising biometric traits for identification. The gait is defined as “Gait is a particular way or manner of moving on foot”. Human gait has a unique personal characteristic and cyclic in nature [3]. The main advantage of gait is that no human interference is required. It can work well even though the camera situated at large distance. In this paper gait recognition method is based at an intersecting points and compare fuzzy and nearest neighbor classifier to achieve better recognition rate. This paper is an enhanced version of gait recognition with shoulder body joint[4].This paper is organized as follows. Section II described the existing method with their utility. Human gait and fuzzy concept described in section III. Section IV provides our proposed method. Experimental results & analysis are presented in section V, followed by Conclusions & future scope in Section VI. II. RELATED WORK Many authors can worked on silhouettes images and Gait Energy images [5], [6], [7]. Haiping Lu et al. [8], [9] defined the body part 22 human parameter. The parameter defined the part length, width position and joint angles. C. Manuscript received January 15, 2013; revised March 22, 2013. Jyoti Bharti is with the Department of Comp.Sc and Engg. of Maulana Azad National Institute of Technology, Bhopal,India, (e-mail: jyoti2202@gmail.com) M. K. Gupta is with the Electronics & Communication Department of Maulana Azad National Institute of Technology, Bhopal, India (e-mail: gupta4@yahoo.com) Senanyake et al. [10] used thigh, shank, foot, hip, ankle and knee body joint for normal gait sequences achieved better result on body joint. Faezeh Tafazzoli[11] extracted the gait motion of leg and arm. They computed the rotation variation of hip, knee and arm of silhouettes images and found better result with arm as compare to without arm variation. It is found from the related works, body joint plays an important role for identification and achieved better result on it. III. HUMAN GAIT AND FUZZY LOGIC A. Human Gait Human identification through a style of walking Style of walking is called gait. Gait can be defined as "A peculiar way or manner one walks". Walking pattern can be analyzed by a gait cycle. Gait cycle is repeated motion of body part[6],[9]. The style of walking or gait cycle of every person is unique[5],[8]. Mostly there are no much more changes in head and shoulder motion as compare to hand and legs. The human gait cycle becomes when the initial position of foot becomes final position shown in Fig. 1. The single gait cycle is further divided into two phases: Stance Phase: Foot touches the ground, loading response, Mid-stance, Terminal stance, and Pre-swing [5], [12]. Swing Phase: Foot does not contact with the ground. Initial swing, Middle swing, and Terminal swing [5], [12]. Fig. 1. Representation of subject moving through the gait cycle. B. Fuzzy Logic Fuzzy logic is mathematical approach to distinguish or classify the object based on the extracted feature values. In a classical way, an object takes on a value of either 0 or 1. Fuzzy logic statements values are lies between 0 & 1. If the value is nearer to 0 then cannot recognize the person and value is closer to 1 then person is identified [13]. There are many types of membership function choose one of them dependent on the application. After testing and survey it decided that trapezoidal function is more suitable for our Gait Recognition with Fuzzy Classification Using Shoulder Body Joint Jyoti Bharti and M. K. Gupta 590 International Journal of Future Computer and Communication, Vol. 2, No. 6, December 2013 DOI: 10.7763/IJFCC.2013.V2.233