Channel Coding And Image Recovery Research Base On Fractal Image Code

With the rapid development of modern communication technology and computer technology, the gross of data is increasing by a geometric seriey speed every day. At the same time , the requirement of communication quality and data rate also enhance continuously. As a result of noise and feeble character in communication channel, the communication signal will fuzzy inescapability when it sent to sink by communication system. In order to insure the correcty of data, error detecting and data correcting are usually used to resolve the problem of data lose in data transmission which is caused by communication channel noise. Fractal compression coding research became a hotspot in technological field due to its high compression ratio. The correcty of fractal compression coding data transmission is an aspect of them.In this thesis, channel coding is used to decrease the error in decompressed image which is transmitted by communication system, according to the character of fractal compression coding data. The contributions are as follows.Firstly, there are several different parts in fractal compression coding. Each part is different in the sensitivity to the communication channel noise, the different sensitivities of each part of fractal compression coding data to the communication channel noise have been analyzed first. Based on it, we design a channel coding of fractal compression coding. By using this method the sensitivity of fractal compression coding data to the communication channel noise has been reduced and the quality of the decompressed image has been improved greatly.Secondly, this thesis proposes a method of checking error blocks caused by communication channel noise and concealing error blocks. This method is good at checking error block with low fault. Then, using genetic algorithm to conceal the error blocks so the image quality is improved. Compared with exhausted searching method, the searching space can be greatly reduced by using genetic algorithm and good image quality is achieved.