A new fingerprint translation finding algorithm

Judging the translation and rotation between the input fingerprint and the template fingerprint is the prerequisite for fingerprint matching, and the process is rather time consuming. We propose a new algorithm of finding the translation between two point sets and its application in fingerprint minutiae matching. This new algorithm is based on the frequency statistic character of the translation vector matrix of two point sets, and has fast processing speed and independence of the locating accuracy of reference points. This method is not only applicable in judging two identical point sets, but also valid in judging two point sets with partial overlapped regions. By using this technique, we achieve a fast speed in finding the translation between two matched minutiae sets, much faster than the traditional method.