Research On Lossless Image Compression With Adaptive Prediction And Statistical Modeling

With the era of Big Data and Cloud Computing coming, Internet data spurt gush, and showing a trend of explosive. In recent years, the use of image data compression m the field of data transmission and storage has got more and more attention, such as biomedical science, fingerprint identification, remote sensing and other aspects of the application. The application of in-depth, image data compression receives more and more attention in the field of data transmission and storage.Image compression is an important field in the research of coding. Due to the adjacent pixels in the image in the time domain, space domain exist a large amount of redundancy and the redundancy in the decoded can also no distortion of the recovered. Thus, it is possible to realize lossless compression of images by using predictive coding.This paper presents a use to be prior experience in image coding to improve the coding system compression performance of adaptive predictive coding algorithm. The algorithm using the preset threshold optimization selected training set. After the training set selected sample points of relevance ranking, the high correlation between the sample points placed in position with larger weights, the weak correlation between the sample points put in the importance of some poor position. Then, using the training set to train the predictor coefficients, and the minimum mean square error criterion to determine the predictor of optimal order. Finally, source probability model is used to initialize the generalized Gaussian distribution, thus by using the optimized predictor to lossless compression of images.Finally, the use of 512* 512 grayscale image using the standard C language in Visual Studio 2010 development platform programming algorithm implementation, verification article image compression algorithm, and associated algorithms for comparison, a higher compression ratio to prove the effectiveness of the algorithm.