The Design And Implementation Of Two Digital Seal Authentication Algorithms Based On Information Hiding Technique

In the fast-developing Internet age, digital office has been very popular. Digital document is an ordinary thing people are faced with every day. However, it is easy for digital documents to be tampered and forged. The problem has aroused more and more attention.The security of digital documents should not be ignored, especially the security of official documents. If attackers succeed in tampering or forging an official document, the consequences will be unimaginable. When an official document is published, the organization or department will usually attach a digital seal to the end of the document. Therefore, we did a lot of studies and analyses of the techniques about digital seal. Then we proposed two digital seal authentication algorithms based on information hiding technique.The first digital seal authentication algorithm is based on index images. The algorithm makes use of the extra space of the palette of a binary image to hide information. The advantage of the algorithm is that it is able to provide a very big embedding capacity. If a binary image is stored in GIF or PNG-8 format, its bpp(bit per pixel) is 7. What’s more, the algorithm will not change the color or position of any pixel. So the algorithm achieves perfect transparency.The second digital seal authentication algorithm is based on bitmap images. The algorithm embeds information by moving the circle’s pixels. Specifically, the algorithm encodes the sequence of a circle’s y-coordinates using RLE(Run Length Encoding). Further, the y-coordinates’ RLE is replaced with a simpler encoding, RLS(Run Length Sequence). During the process of embedding information, the pixels of the first run will be moved to the second run, the pixels of the second run will be moved to the third run and so on. The number of the moved pixels every time depends on the specific information. After the last bit of the information is hidden, the pixels are no longer moved. In the end, the algorithm will generate a new RLS. In the extraction process, we can get hidden information by comparing new RLS with old RLS. The algorithm achieves big embedding capacity and good transparency.Theoretical analyses and experiments demonstrate that the two algorithms can be used to authenticate digital documents with digital seals for practical purpose.