一种基于边缘检测和正负量化的盲水印算法

The Blind Watermarking Algorithms Based on Edge Detection and Quantify the Positive and Negative

提出了一种基于DCT域边缘检测和正负量化的盲水印算法:先用Arnold置乱方法对水印图像进行加密,并对其进行正负变换,然后用边缘检测算子对载体图像进行边缘提取,再根据阈值的设定选择不同的量化值将水印嵌入,提取水印时不需要原始图像的参与,实现了盲提取.实验结果表明:该算法对于裁剪、压缩、噪声、滤波、缩放等攻击都具有较好的鲁棒性和不可见性.

A blind watermarking algorithm based on DCT domain edge detection and positive and negative quantita- tive is proposed. First the algorithm used Arnold scrambling method to encrypt watermark image, and carried on the positive and negative transformation, then used edge detection operator to carrier image edge extraction. Select different quantization value in accordance with the setting of the threshold value to embed watermark. Extract the watermark does not require the algorithm has good the participation of the original image, realizing blind extraction. The experiments show that robustness and invisibility to attacks such as cutting, compression,noise, filtering, scaling