Improving the Viola-Jones Algorithm in Thermal Face Recognition of Images

Abstract

In this paper, a method to improve the performance of Viola-Jones algorithm in finding the relevant parts of a face in thermal images is proposed. First, thermal images are acquired in a semi-controlled environment then images containing noise are cleaned. In the next step, the face image is segmented and then enhanced with linguistic modifiers and DIP techniques. Finally, the Viola-Jones algorithm is applied to detect the parts of the face such as the mouth, nose, and eyes. The experimental results show that the proposed method is effective since it achieves a high hit rate.