FACE PROFILE RECOGNITION AND IDENTIFICATION

Abstract

Since the early 1990's face Recognition Technology (FRT) became anactive research area. Most of the current profile recognition algorithmsdepend on the correct detection of fiducial points and the determination ofrelationships among these fiducial points. Unfortunately, some featuressuch as concave nose, lips, flat chin, etc. make detection of such pointsdifficult and unreliable. Also the number and position of fiducial pointsvary when pose changes even for the same person. In this paper, acurvature-proposed technique is presented, which does not require theextraction of all the fiducial points, but uses information contained in theprofile. The nearest neighbor interpolation method is used to smooth theprofile and then the curvature of the interpolated profile is computed.Using the curvature coefficient values, the fiducial points, such as nasion,chin, and forehead can be reliably extracted using a fast and simplemethod. Then an Euclidean distance method is applied to match the faceprofile based on the curvature coefficient values.Experiments are performed on collected 50 clients with different ages, inpublic area. Each had three shots differences in time capture andilluminations, getting total of 150 images database, recognition rate of96.67% and conclusion are presented and discussed.