Research on Method of Fingerprint Mosaicking Based on Phase Correlation

A method of fingerprint masaicking was advanced according to the image invariant property of rotation,translation and distance.The composite image generated by mosaicking increased the numbers of effective feature.The method included three steps: media filter and a novel normalization based on image local property as image preprocess were deployed to get rid of noise and the frequency deteriorated the accuracy of phase correlation;the images were registered based on phase correlation after they were transformed to frequency domain by pseudo polar fast fourier transform instead of polar fourier transform;the images were mosaicked smoothly.It is proved by a great deal of experiment results that the algorithm can increase the feature numbers and improve the performance of verification.At the same time,the low complexity of algorithm can realize fast mosaicking of fingerprint images.