Fingerprint Alignment Using Similarity Histogram

The performance of fingerprint matching algorithm relies heavily on the accuracy of fingerprint alignment. Falsely aligning two feature sets extracted from two finger images of a fingerprint will increase the false rejection rate (FRR). In order to improve the performance of fingerprint matching algorithm, we present a new fingerprint alignment algorithm called similarity histogram approach (SHA). First, we calculate the local similarity matrix based on minutiae and associate ridges between two fingerprints. Then, similarity histograms of transformation parameters are constructed from local similarity matrix. In the end, the optimal transformation parameters are obtained using a statistical method. Experimental results on FVC databases show that our method is effective and reliable.

• Type: Conference
• Year: 2003
• Where: AVBPA
• Authors: Tanghui Zhang, Jie Tian, Yuliang He, Xin Yang
• Keywords: AVBPA 2003, Fingerprint Alignment, Fingerprint Matching Algorithm, Local Similarity Matrix