Abstract
This paper suggests a new method of recognizing a facial orientation based on tracking a pair of nostrils. Due to the fact that nostril shape is stable compared with other facial parts such as eyes or mouth, the nostril image was treated as a feature point and it was taken in real-time for image processing via an USB camera. When a man faced to up and downward, the darkness areas of both nostrils were increased and decreased, respectively. On the other hand, the difference between both nostril areas could be caused in cases where the face was turn to the left or right directions. The obtained characteristics were utilized for recognition of a facial orientation. Furthermore, the variable threshold in the binarization procedure was applied to enhance the recognition system against the changes in the environmental brightness.
