Matching Composite Sketches to Facial Photos using Component-based Approach

Abstract

Face recognition has wide potential applications in many computer area like face identification, face verification, criminal investigation. The face images can be represented by various representation approaches. Common approaches for representations includes component- based, analytic, patch-based and holistic. Also various researches on sketch recognition which focused on matching sketches drawn by professional artist based on a verbal description of the subject's appearances provided by an eyewitness. In the proposed system composite sketches are synthesized using one of the several facial composite software systems which are publically available. We propose a component-based representation (CBR) approach to measure the similarity score between a composite sketch and mugshot photograph. Using an active shape model (ASM), first automatically detect facial landmarks in composite sketches and face photos. In propose system instead of comparing whole image at the same time, facial components from dataset will be compare for matching. Depending on the component matching images will be display in sorted order. Keywords-Component-based face representation, composite sketch, faces recognition, forensic sketch, heterogeneous face recognition, modality gap.