Research Of Image Encryption Based On LZW Compression Algorithm

Along with the rapid development of computer technology, digital images havebecome a kind of important information carriers. A lot of digital images are produced allthe time in every corner of our life. At the same time, the Internet has changed people’slife; people communicate with each other and share information on the Internet.However, the volume of digital images is very large, and many of them containindividual privacy or include confidential content. So image compression and imageencryption are two important problems which need to be taken into seriousconsideration.In order to compress images, many good compression algorithms can be used, suchas Huffman coding, LZW coding, arithmetic coding and so on. Generally, compressionand encryption are taken as two separate procedures. But it will waste a lot of time ifyou do so. One way to solve this problem is to make the two procedures become one.As a general rule, scholars often introduce encryption into compression algorithms, anda lot of work has been done in this domain. Graphics Interchange Format (GIF) usesLZW coding to compress its image data, and it occupies less spaces and can bedownloaded in a shorter time. What’s more, the partial download of this kind of imagescontain useful information. Thus it is widely used on the Web. The research of GIFimage security is valuable for scientific research and applications in this background.In this paper, two methods to introduce encryption into LZW based on the GIFimages are proposed. In one of the two methods, the dictionary of the LZW is permuted,and the values of color table are changed too. This way doesn’t impact the compressionrate of the LZW coding, and it has good encryption effect in the respects of key spaceand sensitivity. However, this way has a disadvantage, i.e. the result image doesn’t havea uniform histogram. In the other method, the structure of the LZW’s dictionary hasbeen changed. It combines the LZW with the dynamic Huffman coding, and the outputsof are similar to the Huffman coding. This method not only gets very good encryptioneffects, but also improves the compression rate. As the incorporation of compressionand encryption will reduce the time of image processing, it will attract more attention inthe future.