| Abstract: | Information retrieving from internet is easy than before. It is convenient to communicate to each other through Internet. A user can download, store, and send information via Internet. Thus, many problems were appeared such as information security and copyright protection. Many data hiding schemes have been presented in recent years. Information hiding is to embed secret data into cover media for secret data delivering. Because information hiding technique tiny modifies the content of cover medium for implying secret data, unexpected user will not pay too much attention on stego medium. Here, the cover media can be image, audio,video, Webpage, and Execution file.In this thesis, image and web page were used as cover media for information hiding techniques. First Block Truncation Coding, BTC concept is adopted to generate prediction value. Then prediction error is generated by calculating the difference between the prediction value and pixel value. After that the error histogram is generated by counting the prediction errors. Secret data is embedded into cover image by adopting the histogram modification. Second, because the embedding capacity of histogram modification data hiding is highly related to the height of peak point we use difference segmentation strategy to narrow down the range of prediction error. According to the proposed embedding rules theembedding capacity is improved.Third, we use web page as cover medium to deliver secret information. The benefits of using webpage as cover media is easy to deliver secret, because any kind of mobile devices can easy browsing the stego webpage, and the secret can extracted through parsing the source code of stego webpage. Secret data embedding for HTML file is to modify the source code of webpage. Fro English sentence, the sentence is composed by words and blank characters. Also, the blank characters in HGML file can be replaced by special code for implying secret data. For improving the performance of embedding capacity, Cartesian production is adopted in the proposed method. We collect the special blank codes and pairing them by using Cartesian Production. The cover article is partition into nonoverlapping segments. Then, the blank character in a segment will be replaced by special blank code for implying secret data. Experimental results show that the proposed method has better performance then Lee and Tsai’s method in terms of embedding capacity. |
