A Robust Edge Detection Algorithm for Finger Vein Recognition

By extracting vein patterns under the skin of fingers for identifying the authentication, finger vein recognition is a promising biometric recognition technology. A particularly critical step in locating the region of interest of the finger vein image is the edge detection of the finger. Aiming at the problem that the existing algorithm has low detection accuracy on low-quality finger vein images, this paper proposes a robust edge detection algorithm. Firstly, the upper and lower edges of the horizontally placed fingers are detected by using the upper and lower convolution kernels, and then the obtained edges are subjected to erroneous edge judgment and filtering denoising. Finally, edge fitting is performed by BP neural network to repair the wrong finger boundary. The experimental results show that the proposed algorithm can robustly detect the finger edge of low-quality finger vein images.