Description

SOCOVIFI is a research project supported by the Spanish Ministerio de Educación (ref. TIN2006-00829), and by the Andalusian Department of Innovación, Ciencia y Empresa (ref. TIC-1619), including EDRF fundings.
The goal of this project is to design a computer-aided automatic method to assist the forensic anthropology in the identification task by means of craniofacial superimposition. This method will be based on the use of soft computing techniques.

Our proposal focuses on the next three specific targets:

• Design of range image registration techniques based on evolutionary algorithms and metaheuristicis for the reconstruction of 3D skull models and other forensic anthropology objects. The 3D reconstruction is done using partial views acquired by the Konica-Minolta 3D Laser Scanner of the Physic Anthropology Lab at the University of Granada.
• Proposal of automatic techniques for the overlay of the 3D skull model over the 2D face photograph of the missing person. These techniques will be based on evolutionary algorithms and metaheuristics. They will make use of the heuristic information from the anthropometric characteristics of both images to guide the search. Besides, they will consider the inherent uncertainty of the procedure (related to the uncertain location of the landmarks and the partial matching between cranial and facial landmarks) in order to obtain more accurate overlays.
• Development of a fuzzy decision support system to assist the forensic anthropologist in the identification decision making process using the information about the matching between craniometric and cephalometric landmarks. The uncertainty related to the location of the landmarks and the partial matching situation will again be taken into account by considering the different truth degrees in the final decision and the aggregation of different sources of information.Example of automatic skull-face overlay process by means of evolutionary algorithms and fuzzy landmarks:

Example of automatic skull-face overlay process by means of evolutionary algorithms and fuzzy landmarks: