High Efficient Compress Algorithm For Remote Sensing Image

With the development of the Remote Sensing technology,the Remote Sensing images show some trends of "3-H"-High spatial resolution,High spectral resolution,High time resolution,and "3-M"-Multi-platform,Multi-sensor,Multi-perspective.Therefor leads to a huge growth of the data size of RS image.In order to efficiently process,store,and transfer the data,it’s necessary to build some compression algorithm for RS image.It’s meaningful to study how to improve the compress performance,accelerate the compress efficiency,while maintaining the image quality in some Remote Sensing application.Till now,the HEVC is the latest coding standard.It’s a milestone cause the still image compress profile show huge coding performance improvement compared to the widely used image coding standard JPEG.It can achieve a tremendous bit rate saving under the same compress quality,thus it’s optimal for the massive RS data compression.However,as the HEVC standard introduced some new coding tools such as the recursive quad-tree coding structure,which also lead to some extra computational burden,make the time consuming not so welcome in some real application.Thus it’s a key problem to optimize the HEVC framework,to accelerate the algorithm and release some computational burden.This paper presents a research about high efficient compress algorithm for RS image.We first extract some region with special texture,treat it as Side Information.And we proposed a brand-new SI compress algorithm,based on graph path coding.Then with th coded SI,we do some optimization to the HEVC framework,propose a two-parts fast algorithm.We accelerate the CTU partition and prediction mode selection procedure,skip some redundant calculation,improve the compression speed.Finally,the experiment results show that the proposed high efficient compress algorithm can save a lot of time consumption while maintaining the compress quality.