Robust Line Hausdorff Distance for Face Recognition

Human face recognition has a lot of applications in recent life. For last 20 years, face recognition has been an attractive problem for pattern recognition researcher which a lot of method has been introduced. However, non-ideal conditions (lighting conditions, face expressions) for face recognition have been still a significant challenge to researchers. Line Hausdorff distance (LHD), which used distance between edge map of face images for calculating the different between facial images, was a fast and high accuracy method for face recognition. In this paper, a modification of LHD was proposed, Robust Line Hausdorff distance. The experimental results showed that the proposed method had a half of time of the LHD method while the accuracy of face recognition was higher than LHD method. Moreover, the accuracy of the proposed method was robust which the non-ideal condition of face images.