New robust Hausdorff distance-based face detection

This paper proposes a novel approach for robust face detection based on our enhanced Hausdorff distance (HD). A major aim of this research is to achieve a highly efficient method in face detection that can be used in real-time applications. In our approach, a technique to automatically determine an appropriate size of an elliptical model is also introduced. Therefore, there is no requirement for an expert user to intervene in the process. In addition, the enhanced HD proposed in our method consumes much less computation time but provides higher efficiency and more reliability than the conventional HD.