Digital Image Watermarking Based on Normalization and Visual Feature

In this paper, a Discrete Cosine Transform (DCT) domain robust digital watermarking algorithm based on image normalization of invariant moment and the sense of visual feature of person is proposed. In order to boost up the safety and the solidity of the watermark, the watermark is scrambled and chaotically encrypted before embedded into the original image. Firstly, by using image normalization and the invariant centroid theory we can obtain a significant region from the normalized image. Then the significant region is cut into blocks of 8*8, and DCT is performed on the image blocks. While making sure of the watermark's embedding position and strength factor, this paper takes good use of the sense of visual feature of person and the characteristic of image. Experimental results show that the proposed scheme is invisible and robust against attacks.