Eye detection algorithm base on area-blocks pairing and multi-angle template matching

A new approach of eye locating from facial images based on area-blocks pairing and multi-angle template is proposed. In this paper, we first get the face region using skin-color model and extract the edge of face region, then obtain the blocks of suspected eye areas by morphological processing. Last, we can locate the position of eyes accurately using area-blocks pairing and multi-angle template matching. This approach is steady for image-whirling. The experimental results have shown the effectiveness and the robustness of our proposed algorithm.