A scheme for minutiae scoring and its application to fingerprint matching

Minutiae based method is popular in fingerprint matching, however, minutiae are influenced heavily by fingerprint processing and image quality. This paper discusses a scheme for minutiae scoring and its application for fingerprint matching. First, quality factors of three levels, namely global factors, block image quality and neighborhood detailed structure, are investigated and combined by product rule and linear weight, which produce a comprehensive score for minutiae evaluation. Second, minutiae score is incorporated into fingerprint matching by checking the scores of corresponding minutiae pairs. A comparative experiment is conducted on a public fingerprint database, FVC, DB3, and the result presents that minutiae scoring is useful for fingerprint matching.