The Parallel Research On Digital Image Compression Algorithm

Nowadays, many important scientific technologies depend on the computer technology, such as giant scientific computation, computer simulation, etc. The development of the hardware technology in recent decades brought rapid improvement in CPU processing speed, but the processing speed requirements still can not be satisfied in many areas of advanced applications. We need the quick and effective parallel computers to achieve large-scale scientific computing and data processing in many fields, such as SAR image processing energy exploration, weather forecast, military war, artificial intelligence and some others basic researches. The finiteness of single computer's technology determines that multi-computers' parallel will be the direction of the computer development. This also leads to that parallel computing will be the main computing method in the future, and the development of parallel processing technology will be the main power to impel the construction of national defense and the development of the science and technology. Parallel processing technique deals with many aspects, including parallel architectures, parallel software and parallel algorithm, etc.Image plays an important role in our life. We can see many images anywhere, Maybe we can get about seventy percent of information from our eyes. But most of the image compression algorithms have the features including complicated operations, large computing works, huge quantity of data and standardized types of computing. High-speed digital image processing technique will be one of the most important methods in the scientific research. Because of the characteristic of image dates, the least time and space is the most key. Parallel computing is an effective instrument. Along with the developing of high-powered parallel processing system, image parallel processing technology will provide greater space for improving the velocity of image processing. So the researches on image parallel processing technology are very important and valuable.In this paper, the research of the network parallel computing technology, being the main object, is focused on the parallel algorithms in image compression coding applications. Image processing technologies and the basic concept of parallel computing are introduced while showing the parallel image processing technology applications and development. The architecture of parallel computing, parallel computing model and algorithm design methods are paid more attentions. Furthermore we adopted the experiments on a platform MPI to prove the high efficiency of the parallel algorithm. Finally, the feasibility of parallel arithmetic coding was approved by analyzing the characteristics of the sequence arithmetic coding at the same time a model based on LogP parallel arithmetic coding algorithm was derived. Based on the analysis of the complex complexity of time and cost it is convinced that the parallel algorithm proposed in this paper be more high parallel efficiency and well applied to the cluster system MIMD.