Remote Sensing Image Fusion Method Based On Compressed Sensing

Image fusion is an important branch and a research hotspot of the informationfusion. With the rapid development of sensor technology and information technology,So, data is a sharp increase from different satellite sensors.. In order to alleviate theenormous pressure caused by massive data transmission and storage, and can effectivelyextract the images more comprehensive, more accurate and more reliable imageinformation, become a key issue in the field of remote sensing urgent need to address.This caused great pressure for signal sampling, transfer and storage. How to alleviatethe pressure and can effectively extract the useful information bearing in signal havebecame one of the problem urgent need to resolve in signal and information processingfield. The traditional image fusion method due to the great amount of data has beendifficult to meet the need of the actual, Compressive Sensing (CS) provides a new ideato alleviate these pressures.CS is a new kind of sampling theory frame based on information.lt broke throughthe bottleneck of the traditional Nyquist Sampling Theorem, abandon the whole sample,Compressed sensing theory of low sampling rate and compressibility provides a newidea for remote sensing image fusion method.In this paper, the images from Worldview-2satellite, panchromatic image andmultispectral images for the study of remote sensing image fusion. In this paper,combined with CS theory, the main contents as following: Firstly, the traditional fusionhigh sampling rate, high computational complexity, and cannot obtain comprehensiveinformation, compressed sensing theory of low sampling rate, compressible behavioralleviate these pressures provides a new way of thinking. Attention to the signalmeasurement and reconstruction of compressed sensing. Secondly, Traditional pixel-level image fusion method for high-resolution spectral detail is maintained there isa big fusion problem, a figure based on compressed sensing and multi-resolution fusioncutting theory. Spectral minutiae fusion maintained better. Lastly, Based on thetraditional IHS transform fusion methods of compressed sensing theory, on the CS-IHSmethod, it is proposed a block diagonal matrix applied to image fusion, and getobservation matrix image of local information, and then fusions the differentcharacteristics respectively. Compared to sampling random matrix integration program,gets the higher image fusion resolution, be even better edge and texture detail.