Research On Pyramid Parallel Construction Technology Of Massive Remote Sensing Image Data

The traditional stand-alone mode is used to establish a pyramid with a long period of time,and the efficiency is low,and the need to update the map in real time cannot be achieved.Therefore,how to effectively refine the information directly related to the application target from the massive remote sensing image data,and how to effectively process the massive remote sensing data quickly and improve the efficiency of visualization is the most urgent problem to be solved.In recent years,the development of distributed parallel computing technology has provided a new opportunity for remote sensing image processing.This paper mainly studies the parallel computing framework of MapReduce based on Hadoop to optimize the problem that the current traditional single-machine construction technology generates pyramids with low time and low efficiency.Firstly,the pyramid problem is constructed for overlapping and multi-source regions in remote sensing images.The image optimization layering strategy is adopted.Secondly,due to re-sampling As the resolution of the upper middle layer is reduced,this paper uses the improved cat group optimization algorithm to reduce the loss of image resolution,and combines the MapReduce parallel computing framework to construct the pyramid.Finally,the traditional remote sensing image pyramid output technology will have slow output and accuracy.The results show that the cat group optimization algorithm will improve the integrity of remote sensing image data.The efficiency of the parallel pyramid optimization algorithm proposed in this study will follow The number of nodes increases and increases.