Facial Composite System Using Genetic Algorithm

Although humans have excellent facial recognition ability, it is widely recognized that they often have great difficulty recalling facial characteristics in sufficient detail to produce an accurate composite. In this paper, we propose a newly facial composite system - to be coined as genetic-based facial composite system GFCS that depends on genetic algorithm as a facial search method. The system consists of two main activities. In the first activity (database preparation), a number of steps are applied to extract face parts from the database image. While in the second activity (facial composite reconstruction), genetic algorithm is used for reconstructing facial composite image seemed likeness to eye witness facial image. Moreover, GFCS provides utility tools to enhance the resulted facial composite image. The results from using GFCS clarified that the system is successful in producing a good likeness to a witness mental image