The Study On Lossless Compression Algorithm For Ordered Dither Image

The research on efficient compression methods of the ordered dither halftone imagewhich is widely used in practical life and apply for printing and display of continuoustone images, has important application value for its digital store, record and transport.This article introduced the generation, development and application of halftonetechnology, and analysed the current development of halftone image’s compressiontechnology both in domestic and international. As the lossless compression algorithmwas more complexity and it’s coding and decoding time was too long, the bad imagevisual effect after decoding of lossy compression for the ordered dither halftone, twocompression algorithm were proposed on this paper. It’s mainly work as follows:(1) In order to compress the data of dither image with rectangular partition coding,this article designed a preprocessing method based on the dither image’s generationprinciples and characteristic. The dither image has to be down sampling according tothe size of dither template first as well aim at changing it’s original feature that theblack and white pixels are frequently alternated. And as to improve the compressionefficiency, a secondary pretreatment should to be taken on some sampled image.(2) Most algorithms used the inverse halftoning method to get the multiple gray-scale image so as to compress it in a lossy way when compress halftone image withentropy encoding. But the data of image would be inflation after inverse halftoning, andthe quality of decoding image must be unsatisfactory if to compress it in a high rate. Asfor this problem, a lossless compression algorithm based on the huffman coding wasdesigned on this paper that the halftone image should be transformed into gray imagewith the method of binary to decimal conversion by partitioning it with threshold template according to the relevance of the image block. And as to enhance thecompression ratio, the halftone image had to be xor block by block first.At last, in order to prove the feasibility and practicability, the above methods wereprogramming in the Windows XP operating system environment by VC++6.0.