A Study of High Capacity and Reversible Image Information Hiding Schemes

• Main Authors: Lin-Chun Li, 李林軍
• Other Authors: Ching-Te Wang
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/34268991039110337786

碩士 === 國立勤益科技大學 === 研發科技與資訊管理研究所 === 100 === The data communication is becoming more and more rapidly and conveniently due to the information technology and Internet expeditious development. Therefore, the protections of information security and integrity are becoming very important in the network transmission. The data hiding schemes and the embedding security are needed to consider in the image research, but how to enhance embedding capacity without losing image quality is become priority research centers. Spatial domain information hiding using histogram modification techniques to hide secret has become trend, but the lower embedding capacity often subject to the histogram peak point frequency restriction. For this reason, we will propose reversible data hiding techniques based on the neighboring pixels prediction method that using two levels prediction makes the pixels effective exploitation. In the first phase, we use the neighbor pixels, median and mean values to evaluate the prediction difference and compute the difference histogram of the image. The second phase uses an inverse S form to scan image and compute the prediction difference that can generate the second difference histogram. By the use of two histograms, the system generates two pairs of peak point, which can be used to embed the watermark and adjust the hiding capacity. Furthermore, we promote a greater hiding capacity that using pixel frequency of block as prediction data hiding technique. This method explores the frequency of pixels in each block and uses them as the predictive values. Because the neighboring pixels of blocks have the similar characteristics and the differences of prediction pixels are approximately that the hiding capacity can effectively increase. Furthermore, our study will propose the new data hiding technique that design a dynamic matrix to embed secret information. We use two random numbers to generate a dynamic matrix, and the remainder of pixels match with matrix to embed the information. In the hiding process, we use four pixels as one group to embed two of novenary system data, and the capacity is higher than LSB hiding technique. Thus, the proposed data hiding techniques can increase the image carrying data capacity. In this study, each pixel in the image only adjusts one unit, and the stego-image can attain to excellent quality. Moreover, our data hiding techniques are always the reversible data hiding, the stego-image is extracted information without use the original image information, and can still be reversed to the original image. From the experimental results, the histogram modification techniques and the dynamic matrix data hiding schemes not only have high capacity and good PSNR values of image quality, but also have reversible characteristics of data hiding.