The Research And Implementation Of Some Key Technologies On High-resolution Remote Sensing Image Sharing

With the development of the aerospace and electronics industries, the quantity of optical satellite sensor technology continues to improve. Thus remote sensing image resolution greatly improved, the capacity of a single image has become increasingly large. In recent years, China’s resource series of satellites have produced mass of high-resolution remote sensing images. It will promote the development of research and application of remote sensing in various industries by the implementation of rapidly and effectively sharing these images. On the high-resolution remote sensing image sharing,traditional way is outdated, fast sharing way in distribute environment will be a big trend.In this paper, some key technologies on the organization&management and transmission of remote sensing images are to research as follows:1. Based on the analysis of three commonly used remote sensing image resampling algorithm, improve a Resampling of Algorithm based on Distance weighting and Linear stretching(RADL). The algorithm are designed to work well accuracy under certain performance.2. On the basis of compare the performance of image slices in quad-tree with in sequential way, develop a quad-tree encoding method matching the sequential way to create image pyramid. Analyze the searching ways of slice images in the encoding method.3. Introduction of CUDA technology, improve the computing model of image re-sampling algorithm. Based on the analysis of time-consuming, develop a fast image calcution mode on threshold selection to create image pyramid.4. Build a P2 P model transfer system and design load balancing mechanism for service&client cluster. Develop prototype system combing the specific industry application demand.Finally, RADL algorithm achieves increasing by 9% of the experimental results on PNSR. On rapid image threshold selection computing model of the pyramid creation method, achieves 12 times traditional quadtree computing model and relatively 1.2times to CUDA computing model in speed-up. On image transmission, achieves rate increasing linearly while the number of concurrent requests growing and CPUutilization leveling off. These key technologies, are widely used in remote sensing sharing data platform to provide data&product services in web publishing way.