ABSTRACT
In Current Scenario the increased rates of crime associated in house, road, parking place, ATM Security built up by surveillance techniques have been main agent for industrial and academia. Camera are already install in ATM, road, parking place, public place for capturing real time movement with facial image of users but the function in imitated for criminal investigation which could be useless when a criminal’s face is to be occluded .Occlusion can be glass,mask,scarf, beard etc. also to detect face in unconstrained environment challenging due to various poses, illuminations. For solving the current issue the feature extraction method named as local patterns of gradients (LPOG) is introduced for robust face recognition. Based on LPOG descriptor, we have proposed a novel face recognition system which exploits Whitened Principal Component Analysis (WPCA) for dimension reduction and weighted angle based distance for classification. Experimental results on user define database prove that LPOG+WPCA with SSR Enhancement robust against have a wide range of challenges such as: illumination, poses, occlusion, pose, time-lapse variations and low resolution and classify face with SVM classifier.
