The Research Of Quasi-Lossless Compression Methods And Efficiency For Remote Sensing Image Based On Wavelet Transform

With the implementation of the Digital Earth plan and the development of remote sensing information technology, the remote sensing images, with high-capacities, rich-details and more-textures, set an extremely high request with data storage and transmission. As the remote sensing images are important to analysis and recognition the terrain features, in most cases the remote sensing images are wanted to be compressed effectively and efficiently with the lossless information, it also makes different between remote sensing image compression technology and normal image compression technology. Therefore, seeking for newly and effectively mathematical analysis method to replace traditional analysis method is becoming the crucial factor of remote sensing image compression development. According, this paper proposed the remote sensing image quasi-lossless compression design based on wavelet transform, and the results prove better.At first, the paper reviews image compression coding technology and its theoretical foundation, introduces the evaluation criteria and indicator of image quality. Secondly, it explains the historical development and basic concepts of wavelet transform, discusses several image coding algorithms based on wavelet, analyzes and appraises the application that wavelet transform using in the image compression coding, it provides the theory basis for the mapping transform and entropy coding of remote sensing image compression design. Then, it analyzes the constitution and correlation of remote sensing image, and the coefficients which decomposed after the wavelet transform, it draws the conclusion that it couldn't solve the efficiency of remote sensing image compression only using the method of mapping transform and entropy coding, it provides the theory basis for the quantization of remote sensing image compression design. At last, the paper proposes the remote sensing image quasi-lossless compression design, which includes the mapping transform module based on lifting integer wavelet, the quantization module based on texture frequency division, the entropy coding module based on optimized truncation embedded coding. It explains the realization of above three module process in detail, the results get from MATLAB simulation are compared with the results get from the method of JPEG2000 lossless compression. According to the experimental results, under the premise that the remote sensing image quality doesn't decline, this quasi-lossless compression design improves the performance and the efficiency of remote sensing image compression, it also meets the request of remote sensing image compression.