An Intelligent Multi objective Progressive Algorithm for Identification of Surgically modified Faces

Abstract- Evolution and low-cost leads to the popularity of plastic surgery techniques and it considerably changes facial aspect and alterations. Social constraint and decreased costs are stimulating more and more people to undergo plastic surgery techniques for changing feature flaws like aging, removing birth marks and disguise, which will lead to a serious challenge for face identifying algorithm. As a result, indentifying individuals after plastic surgery procedures is becoming a relevant problem for law-enforcement agencies to identify face before and after plastic surgeries. In the Current state of the art face recognition algorithms, Extended uniform circular local binary pattern based descriptors, Fast nearest-neighbour algorithm, Granular methods were used. But these methods have trouble like Limited variability in the self-report scores, Multiple Atlases, and Inversion effects. So a new system is proposed with the following phases, (i) Pre-processing of images from data set, (ii) Active Shape Model(ASM) for Feature extraction using Localization of face and primary facial, (iii) Multiple Instances Support Vector machine (MI-SVM) for Classification of feature subset, the above listed modules will refine the accuracy and robustness of the Facial identification system