Spatially eigen-weighted Hausdorff distances for human face recognition

✍ Scribed by Kwan-Ho Lin; Kin-Man Lam; Wan-Chi Siu

• Book ID: 104161678
• Publisher: Elsevier Science
• Year: 2003
• Tongue: English
• Weight: 325 KB
• Volume: 36
• Category: Article
• ISSN: 0031-3203
• DOI: 10.1016/s0031-3203(03)00011-6

✦ Synopsis

Hausdor distance is an e cient measure of the similarity of two point sets. In this paper, we propose a new spatially weighted Hausdor distance measure for human face recognition. The weighting function used in the Hausdor distance measure is based on an eigenface, which has a large value at locations of importance facial features and can re ect the face structure more e ectively. Two modiÿed Hausdor distances, namely, "spatially eigen-weighted Hausdor distance" (SEWHD) and "spatially eigen-weighted 'doubly' Hausdor distance" (SEW2HD) are proposed, which incorporate the information about the location of important facial features such as eyes, mouth, and face contour so that distances at those regions will be emphasized. Experimental results based on a combination of the ORL, MIT, and Yale face databases show that SEW2HD can achieve recognition rates of 83%, 90% and 92% for the ÿrst one, the ÿrst three and the ÿrst ÿve likely matched faces, respectively, while the corresponding recognition rates of SEWHD are 80%, 83% and 88%, respectively.