关键词: 图像水印，奇异值分解，虚警错误

Abstract: The reasons of high falscpositive detection rate problem for a class of singular value decomposition(SVD)based image watermarking algorithms were analyzed and the corresponding results were given. On this basis, a novel watermarking algorithm based on discrete wavelet transform(DWT) and SVDwas proposed. First the host image is divided into non-overlap blocks and each block is applied 1-level DW T, and then SVD on the low-frequency sub-bandis made. The watermarking is preprocessed by Arnold transform and then is embedded into the singular values of the low-frequency sub-band based on quantization method. Experiments show that the proposed algorithm can overcome the high falscpositive detection rate problem which exists in the traditional watermarking algorithm based on SVD and robust against common signal processing such as JPEG compression, filtering, and the adding of noise.

Key words: Image watermarking,Singular value decomposition,False alarm errors