3D Face Recognition using Horizontal and Vertical Marked Strips

Abstract

This paper presents a hybrid 3D face recognition method based on point cloud conversion method and horizontal and vertical stripes calculation method. For this method, first of all 3D face model is converted to point cloud. Then the face region is extracted from forehead to lips in the image by finding the nose tip. Next, the face region is normalized. After normalization, the face image is divided into horizontal and vertical strips which are used to calculate average rows and columns values. Then average values of rows and columns are taken and compared, as a result of which the minimum resulting average value is matched with the database. The proposed method has been applied on the GAVAB database. The outcome indicates the best classification with rows and columns values and also provides good recognition results over the previous similar methods.