User identification is a very interesting andcomplex task. Invasive biometrics is based on traitsuniqueness and immutability over time. In forensic field,fingerprints have always been considered an essentialelement for personal recognition. The traditional issue isfocused on full fingerprint images matching. In this paper anadvanced technique for personal recognition based onpartial fingerprint is proposed. This system is based onfingerprint local analysis and micro-features, endpoints andbifurcations, extraction. The proposed approach starts fromminutiae extraction from a partial fingerprint image andends with the final matching score between fingerprint pairs.The computation of likelihood ratios in fingerprintidentification is computed by trying every possibleoverlapping of the partial image with complete image. Thefirst experimental results conducted on the PolyU (HongKong Polytechnic University) free database show anencouraging performance in terms of identificationaccuracy.
|Number of pages||7|
|Publication status||Published - 2013|
All Science Journal Classification (ASJC) codes
- Artificial Intelligence