Forensic sketch based face recognition using geometrical face model

Abstract

Face recognition using sketches plays a major role in the field of forensic investigation and the field is considered to be more challenging domain for research work. The sketch forms a visual representation of a face and it can be utilized in many recognition applications for instance in facial expression recognition, retrieval of face, in the field of law enforcement and sketch-to-photo recognition. In this paper, a sketch based face recognition system is designed and processed with three main phases namely face detection to locate the face region in the input image using AdaBoost algorithm. In the next phase, important components which constitute face structure are extracted using geometrical model of a face. Then from each facial component region, a texture local descriptor called Weber Local descriptor features are extracted to obtain significant properties of face. Finally, each input image is compared with the database and classifies the sketches using ANN classifier. The performance of our proposed method is compared with existing methods and proves to be a better approach.