A Novel Probabilistic Image Authentication Methodbased on Universal Hash Function for Rgb Images

摘要

In this paper, a novel hash based image authentication for RGB (Red, Green, Blue) images is proposed. Firstly, universal hash function is modified according to R, G and B layers and 3 different function is obtained to watermarking generation then, watermark generation is performed by using these equations. In this method, one layer is used for watermark embedding. Modulo based data hiding method is used for watermark embedding. To provide confidentiality of the proposed method, chaotic random number generator that is logistic map is used. The embedding layer and watermark generation function are selected by using logistic map. In watermark extraction, modulo operator is used. Watermark generation step is repeated for tamper detection. The extracted watermark and generated watermark are compared with each other. If these watermarks are equal, image is authenticated, otherwise tamper detection is detected. In the proposed method, DWT (Discrete Wavelet Transform) and SVD (Singular Value Decomposition) are used for generating perceptual hash of images and recovering images. Experimental result showed that the proposed method achieved high image authentication ability with minimum distortion.