Partial Fingerprint Matching via Phase-Only Correlation and Deep Convolutional Neural Network

A major approach for fingerprint matching today is based on minutiae. However, due to the lack of minutiae, their accuracy degrades significantly for partial-to-partial matching. We propose a novel matching algorithm that makes full use of the distinguishing information in partial fingerprint images. Our model employs the Phase-Only Correlation (POC) function to coarsely assign two fingerprints. Then we use a deep convolutional neural network (CNN) with spatial pyramid pooling to measure the similarity of the overlap areas. Experiments indicate that our algorithm has an excellent performance.