Some Information Hiding and Compression Methods based on Side Match Vector Quantization

• Main Authors: Jia-Ping Chen, 陳家苹
• Other Authors: Kuo-Lung Hung
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/68760995604650783879

碩士 === 朝陽科技大學 === 資訊管理系碩士班 === 100 === This paper presents two simple and effective methods of reversible information hiding. One is for the general reversible information hiding and the other is for reversible hiding method under the VQ encoded which is also used for image compression. These two methods are both based on the SMVQ technology. The first method is a reversible information hiding technology based on SMVQ. In the method, the whole image is first divided into complex blocks and smooth blocks by using SMVQ coding technology. We put these two different types of blocks of pixels to be predicted using different methods. After all the pixels of the prediction are computed, then the difference between the conduct is calculated to embed the secret information. The second method, we propose a new image compression technology which is also based on SMVQ. Firstly, the codebook is sorted by using PCA algorithm, and then the codebook is used to predict for SMVQ coding purposes. After obtaining SMVQ index table, the index difference is then calculated between VQ and SMVQ. The secret information will be embedded into index difference using Huffman compression coding. Experimental results show that our proposed methods can effectively increase the secret information embedding capacity, and can also improve the image quality for human vision range. Using SMVQ predictive coding technology not only can reduce the image bit but also increase the image compression rate.