Abstract–In this paper a contour matching based face recognition system is proposed, which uses “contour” for identification of faces. The feasibility of using contour matching for human face identification is presented through experimental investigation. The advantage of using contour matching is that the structure of the face is strongly represented in its description along with its algorithmic and computational simplicity that makes it suitable for hardware implementation. The input contour is matched with registered contour using simple matching algorithms. The proposed algorithm is tested on BioID face database and % recognition rate is found to be 100%. The proposed system of face recognition may be applied in identification systems, document control and access control. Index Terms—Face and gesture recognition, image processing and computer vision, pattern analysis, pattern recognition I. INTRODUCTION Face recognition [1,2,3] is a form of biometric identification. A biometrics is “Automated methods of recognizing an individual based on their unique physical or behavioral characteristics.” The process of facial recognition involves automated methods to determine identity, using facial features as essential elements of distinction. The automated methods of facial recognition, even though work very well, do not recognize subjects in the same manner as a human brain. The way we interact with other people is firmly based on our ability to recognize them. One of the main aspects of face identification is its robustness. Least obtrusive of all biometric measures a face recognition system would allow a user to be identified by simply walking past a surveillance camera[4,5]. The research on face recognition has been actively going on in the recent years because face recognition spans numerous fields and disciplines. There is an increasing demand for security in commercial and law enforcement applications. The rapid development of face recognition is due to a combination of factors such as active development of algorithms, the availability of large databases of facial images, and a method for evaluating the performance of face recognition algorithms. Manuscript received January 9, 2008. (Write the date on which you submitted your paper for review.) S.T.Gandhe is research scholar pursuing his PhD from Visvesvaraya National Institute of Technology, Nagpur, India (Email: star_stg@yahoo.com) . K.T.Talele is Assistant Professor in Electronics Engineering Department, S. P. College of Engineering (Unaided), Andheri (w), Mumbai. ( Email: kttalele@yahoo.co.uk) Dr.A.G.Keskar is Professor in Electronics and Dean R & D, Visvesvaraya National Institute of Technology , Nagpur, India ( Email: avinashkeskar@yahoo.com ) While humans quickly and easily recognize faces under variable situations or even after several years of separation, the problem of machine face recognition is still a highly challenging task in pattern recognition and computer vision. A face is inherently a 3D object illuminated by a variety of lighting sources from different directions and surrounded by arbitrary background objects. Therefore the appearance of a face varies tremendously when projected onto a 2D image. Different pose angles also cause significant changes in 2D appearance. Robust face recognition requires the ability to recognize identity despite such variations in appearance that the face can have in a scene. Simultaneously the system must be robust to typical image acquisition problems such as noise, video camera distortion, and image resolution. The recognition methods are categorized as follows which is based on intensity images [6,7,8,9,10]: Holistic matching methods use the whole face region as the raw input to a recognition system. One of the most widely used representations of the face region is eigen pictures [11,12,13,14,15,16], which are based on Principal Component Analysis (PCA). Using PCA, many face recognition techniques have been developed: eigenfaces, which use a nearest neighbor classifier; feature-line-based methods, which replace the point-to-point distance with the distance between a point and the feature line linking two stored sample points; Fisher faces which use linear/Fisher discriminant analysis (FLD/LDA); Bayesian methods, which use a probabilistic distance metric; and SVM methods, which use a support vector machine as the classifier. Utilizing higher order statistics, independent-component analysis (ICA) is argued to have more representative power than PCA, and hence may provide better recognition performance than PCA. Being able to offer potentially greater generalization through learning, neural networks/learning methods have also been applied to face recognition. One example is the Probabilistic Decision-Based Neural Network (PDBNN) method and the other is the evolution pursuit (EP) method. In Feature-based matching methods, local features such as the eyes, nose, and mouth are first extracted and their locations and local statistics (geometric and/or appearance) are fed into a structural classifier. Earlier methods belong to the Feature-based matching methods, using the width of the head, the distances between the eyes and from the eyes to the mouth, etc, or the distances and angles between eye corners, mouth extrema, nostrils, and chin top. More recently, a mixture-distance based approach using manually extracted distances was reported. Without finding the exact locations of facial features, Hidden Markov Model (HMM) based methods use strips of pixels that cover the forehead, eye, nose, mouth, and chin reported better performance than by using the KL projection coefficients instead of the strips of Face Recognition Using Contour Matching S. T. Gandhe, K. T. Talele, and A.G.Keskar IAENG International Journal of Computer Science, 35:2, IJCS_35_2_06 ______________________________________________________________________________________ (Advance online publication: 20 May 2008)