STEGANOGRAPHY BASED ON SELF-ORGANIZING COMPETITIVE NNS AND HVS
In order to provide larger capacity of hidden secret data and imperceptibility of stego-image, a novel steganographic method based on self-organizing competitive neural networks (NNs) and human vision system (HVS) was presented. First, self-organizing competitive NNs based on competitive learning was trained according to contrast and texture sensitivity, the NNs trained could thus distinguish the pixels in less sensitive areas from those in more sensitive areas. Then the NNs trained was exploited to estimate the degree of sensitivity of pixels so that more secret data could be embedded in less sensitive areas of vision. From the experimental results, compared with the Four-sided Side Match method, the proposed method hides much more information and maintains a good visual quality of stego-image as well.
Steganography HVS Side match Sensitivity Self-organizing competitive NNs
Z.W.Kang J.Liu Y.G.He
School of Computer and Communication, Hunan University, Changsha 410082,China School of Electrical and Information Engineering, Hunan University, Changsha 410082, China
国际会议
杭州
英文
395-397
2006-10-12（万方平台首次上网日期，不代表论文的发表时间）