Sclera Recognition by Density Sampling Features Based Vascular Structure Rapid Matching for Identity Identification

In this study, the effective sclera vascular characteristics identification method is developed to recognize the identity through eye images. Firstly, the proposed design uses the improved Daugman algorithm for the iris fast location. Based on the iris circle, the sclera region is quickly segmented through the color information. Next, the vascular features are enhanced by the adaptive histogram equalization and Gabor filtering process. Then the density samples are processed and the feature points are established. The dense scale-invariant feature transform (Dense-SIFT) based vectors are calculated at the neighborhood in each feature point, and the structural vascular feature matching process is enabled. Finally, the random sample consensus (RANSAC) process is applied to increase the matching efficiency, and the optimal match pair meets the corresponding relationship on the geometry, and then the recognition accuracy is improved effectively. By the dual-core PC at 2.83GHz operational frequency, the average processing time is less than 1 second, and the accuracy rate of the proposed system is up to 96%. Both the false acceptance rate (FAR) and the false rejection rate (FRR) are less than 3%.