A Novel Clonal Selection Algorithm for Face Detection

Abstract

Based on the Antibody Clonal Selection Theory of immunology, we put forward a novel clonal selection algorithm and apply it into face detection problem, named by CSAFD. The new detector is fast and reliable, which can detect faces with different sizes and various poses from both indoor and outdoor scenes. The goal of this research is to detect all regions that may contain faces while remaining a low false positive output rate. The new algorithm firstly abstracts the face template and then realizes the precise location of the face using clonal selection algorithm and template matching. When compared with immune genetic algorithm, standard genetic algorithm and evolutional genetic algorithm, the experimental results on images with moderately complex background scene showed, the new algorithm had a good performance and was much more accurate and robust.