A Similarity Measure Based on Hausdorff Distance for Human Face Recognition

A similarity measure based on Hausdorff distance (SMBHD) for face recognition is proposed in this paper. Different from the conventional Hausdorff distance based measures, the proposed measure can provide not only the dissimilarity information but also the similarity information of two objects to compare them. The added similarity information can especially better the discriminating capability of an object recognition system for similar objects such as faces with variant lighting condition and facial expression. In order to evaluate the performance of a face recognition system using the proposed similarity measure based on Hausdorff distance (SMBHD), the face images included in the AR, ORE, and Yale face databases have been used. The Experimental results show that the system has a better performance than the systems based on conventional Hausdorff distance measures and the eigenfaces approaches