A Memetically Optimized Weighted Based Approach for Matching Sketches with the Digital Face Images
Kushal J Masarkar1, Lilesh P Wankhade2, Rita Dhage3
1Mr. Kushal J Masarkar, Electronics Engg. Department, RTM Nagpur University, KITS Ramtek, India.
2Mr. Lilesh P Wankhade , Electronics Engg. Department, RTM Nagpur University, B.D.C.O.E, Sevagram, India.
3Miss. Rita S Dhage, Digital Communication Engg, Department, RGPV Bhopal University, CIIT Indore, India.
Manuscript received on March 26, 2013. | Revised Manuscript received on April 11, 2013. | Manuscript published on April 30, 2013. | PP: 178-181 | Volume-2, Issue-4, April 2013. | Retrieval Number: D1344042413/2013©BEIESP

Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: One of the important cues in solving crimes and apprehending criminals is matching sketches with digital face images. This paper presents an algorithm that extracts discriminating information from local regions of both sketches and digital face images. All details information present in local facial regions are encoded using multi-scale circular Weber’s local descriptor. We propose a novel discriminative descriptor modified WLD i.e. multi-scale circular Weber Local Descriptor.. It is inspired by Weber’s Law, We organize MWLD features to compute a histogram by encoding both differential excitations and orientations at certain locations of an sketch and digital face image. Further, an evolutionary memetic optimization approach is proposed to assign optimal weights to every local facial region for identification purpose. Foreign sketches drawn by sketch artist is of poor quality, a pre-processing technique is used to enhance the quality of images and improve the identification performance. Comprehensive experimental evaluation on different sketch databases show that MCWLD proposed algorithm yields better identification performance compared with the existing face recognition algorithms.