Face recognition using interpolated Bezier curve based representation

This paper proposes an efficient face recognition scheme, which incorporates certain methods. In this scheme, edges are obtained from sobelled face images. The major feature areas are segmented and principal curves are obtained by applying thinning algorithm on the features of resulting face image. A modified thinning algorithm based on line sweep procedures is used. A line sweep is a process where the plane is divided into parallel slabs by lines passing through certain "events" and then items are processed according to an order of the slabs [A Line Sweep thinning Algorithm]. A new method for Bezier contour point interpolation is used after locating the two end points of the curve on the prime facial features. The two control points of a Bezier curve are interpolated by this method. The Bezier points i.e. the two end points of a curve and the two control points, form the line segments whose Hausdroff distances are calculated as specified in the algorithm discussed in this paper. The recognition rate is 100% on normal faces and also on faces with some simple expressions.