Fingerprint Verification Based on Absolute Distance and Intelligent BPNN

A fingerprint verification system based on absolute distance and intelligent BPNN is proposed. An image- based method with invariant moment features for fingerprint matching is to overcome the demerits of the traditional minutiae-based method and Gabor feature- based method. The fingerprint verification system contains two stages: the off-line stage and the on-line stage. Each stage has a feature extraction model. A total of four sets of seven invariant moment features are extracted from four partition sub-images of a Region of Interest (ROI). Measuring the similarity between the feature vectors of test fingerprint and those of template fingerprint in the database is implemented by two methods, the absolute distance and the Back Propagation Neural Network (BPNN). The experimental results show that the proposed method with BPNN matching has a higher matching accuracy, while with absolute distance matching has a faster matching speed. Comparing with a traditional method, the proposed method performs better with a higher recognition rate, too.