A personal identification system based on isodensity maps in which the properties of a face are well reflected has been proposed. In the conventional method, however, there were problems in terms of the memory requirements for the registration of face images and the processing time. To alleviate these problems, a new matching scheme using chain coded isodensity maps is proposed. Isodensity contours are line pictures. Therefore, they can be converted to chain codes. By using the coded isodensity contours, the improvement of the processing performance can be expected in terms of the processing time and the memory size required. In this paper, a robust personal identification system for angled faces based on the chain coded isodensity maps is proposed. © 2009 Wiley Periodicals, Inc. Electron Comm Jpn, 92(7): 57–65, 2009; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecj.10035...

Fast matching algorithm for isodensity contours of faces based on chain coded description and interpolation of isodensity contours

10.1002/ecj.10035