The Study Of Images Denoising And Fusion Algorithm Of Multi-source Remote Sensing

With the rapid development of remote sensing technology, the new sensors areemerging, a large number of different spectral resolution, spatial resolution image datacan be obtained in the same area, using remote sensing technology. Multi-sourceremote sensing image fusion can take full advantage of these in time, spatial andspectral remote sensing images to improve the utilization of information and enhancedata reliability and availability which can effectively overcome the ambiguity of asingle sensor image data.This paper focused on the study of images denoising and the pixel-level fusion ofmulti-source remote sensing images on the basis of the accurate registration of SPOTimages and TM multi-spectral image. The work of this paper is as follows:(1) introduced the present research situation of image fusion in China and abroad,as well as the level of the image fusion technology development and the evaluation offusion algorithm. On the basis of the analysis of the common noise in remote sensingimage and denoising method, mainly focused on wavelet threshold denoising method,and proposed an improved method based on wavelet half-soft threshold denoisingmethod, through simulation results obtained effective denoising method for imagewith Gaussian noise, salt and pepper noise;(2) Realized the traditional wavelet transform fusion with SPOTand TM images.Introduced the IHS transform in image fusion applications, and Realized thealgorithm which combined IHS transform and Wavelet transform to fuse the imagedata.Simulation results show that the IHS wavelet transform fusion algorithminherited the advantages of the IHS algorithm, the fused image not only have a goodvisual effects, but also has a higher resolution.(3) Focused on the choice of several key factors in the wavelet transform fusionprocess, they were the choice of wavelet basis function, the selection ofdecomposition level and fusion rules. And improved and optimized the wavelettransform fusion method on this basis, and imported a method which is based onmulti-scale wavelet transform fusion method. The experimental results show that thetwo algorithm is effective to enhance the performance of image fusion. (4) Discussed the Curvelet transform image fusion algorithm, and realized theimage fusion algorithm based on Curvelet transform on the Matlab, and at last,through the experimental comparison shows that the algorithm can be good to retainthe spectral information of the source image, while improving the spatial resolution ofthe fused image.