Forensic Facial Image Analysis Providing 3D Mapping, Meta-tagging, Comparative Operation and Search System

With the growing concerns surrounding security and terrorism around the world, biometrics has become one of the premier solutions to combat these issues. By extracting unique anatomical features and markings we can not only learn the identity of subjects but also their associations in social groups based on our novel research regarding Scars, Marks, and Tattoos (SMT). In this proposal, we present a solution that will significantly lower identification time of any face recognition algorithm by reducing the overall number of"possible subjects"through the use of soft biometric indexing and identify potential linkages to groups based on extracted SMT data. Based on our significant prior published experience in this area and current Phase II SBIR research and development (ONR N08-077 Automated Entity Classification in Video Using Soft Biometrics), we will construct a solution that leverages soft biometric features (e.g. age, race, gender, skin, hair, eye color, scars, marks, and tattoos etc.) which can be automatically extracted from face images based on their color, shape, and relation to facial landmarks. Finally, our system will leverage our novel 2D-to-3D facial reconstruction technology to provide accurate results regardless of subject cooperation and facial pose (-90 to 90).