A robust biometric watermark-based authentication scheme

In this paper, we propose a novel technique in which biometric watermark, namely, fingerprint is used as authentication information. To effectively verify an image by fingerprint watermark, we utilize some techniques like Gabor algorithm, locally adaptive thresholding, Hilditch´s thinning method together with heuristic rules, and Hamming measurement to optimally extract minutiae vector from fingerprint as well as improve accuracy of matching process. Furthermore, to make our scheme robust, the lowest frequency of DCT blocks is chosen to hold watermark. With proposed method, the authentication scheme can achieve high capacity and good quality. The experimental results show that our algorithm can be robustness against to JPEG compression, brightness and contract modification, Gaussian noise, adding blur, cropping, rotation, scale, translation, edge detection, emboss, outline, bulge, twist, sharp image.