一种基于Canny边缘检测的隐写算法

iamond encoding (DE) increases the embedding capacity and stego image security through pixel pair embedding. In traditional DE methods, all the pixels of an image conceal confidential information with the same embedding rate, but actually different area can conceal different capacity of confidential information to improve embedding rate. In order to get better stego image, it is divided into contour and non-contour area by canny edge detection and different regions embed information with different embedding rate. The principle of embedding process is that smaller embedding rate is used in contour area and larger in non-contour area. According to the different embedding rate, three methods are presented here, named CANNYDE11, CANNYDE12 and CANNYDE22 method separately. Experiments proves that CANNYDE algorithms outperform other algorithms such as LSB replacement, random LSB matching, the DE algorithm and adaptive edge detection algorithm with the Markov security indicator and SPAM security measure, and they also have a good perceptual effect through PSNR measure.