A Non-domination Pareto-based Scale-Invariant Approach for Face Recognition

Human face recognition has presented a major challenge to the researchers from different domains enabling to enhance the security and pattern analysis. The different orientations, lighting, pose and facial expression of a human face constructs an array of similar images with variations. The identification of face under different circumstances has been the focus of the researchers from a decade to the present time. For face recognition   higher the number of feature may not lead to high recognition rate. Hence, the selection of the optimum features becomes primary concern. It reduces the feature size and increases the recognition rate.   Many algorithms have been proposed that fulfilled the goal of face recognition system but also comprises of some drawbacks. In this paper a novel Pareto-Optimized Evolutionary Approach with Scale Invariance Discrimination has been proposed. The algorithm extracts the set of relevant features from the given image. The optimization of the features is performed for finding the features that enhances the recognition rate. The algorithm performs the classification of the test image, given the set of training images to obtain the accuracy of human face identification. The recognition rate is evaluated to show the performance of the proposed approach with conventional methods.