Accuracy Enhancement Scaling Factor of Viola- Jones Using Genetic Algorithms

摘要

Face detection is considered an important topic in the field of computer vision. Despite the existence of several algorithms that have been used in objects recognition, to the best of our knowledge, Viola-Jones proved itself as a powerful object recognition approach, particularly face recognition. Viola-Jones is governed by a number of parameters that affect its performance such as suppression, scaling factor, search mode, scaling mode, mode of selected item, scaling of selected item and use parallel processing. The scale factor has the most impact on the process of the fact recognition and introduces balance between accuracy and speed in the process of face-detection. In this paper, the performance of Viola-Jones is improved by modification of the scaling factor using genetic algorithm when applying our proposed approach on some general datasets, the results have been promising.