The Research On Removing Thin Cloud Of Satellite Remote Sensing

The satellite remote sensing technology has been developed several years, which has already obtained many encouraging achievements. It is widespread applied day by day in the national economy. The satellite remote sensing image digital processing technology is also more and more advanced, but which still has certain insufficiency. Such as the noise elimination is a topic which is not completely solved.In the satellite remote sensing image, the noise influence is one of important reasons which makes it is used scarcely. In the satellite remote sensing image noises, the cloud cover is one of the most main noises, which not only reduces the interpretation precision of the image, but also brings difficulties for the following process of the image. So how to eliminate the influence of thin cloud is very important.The work in this thesis focuses on removing thin cloud cover in one image.The main research content is followed:(1) Gain the satellite remote sensing image and analysis the fuzzy mechanism.(2) The imaging model and it's restoration method with thin cloud cover.(3) The Fourier Transform and it's nature which process the remote sensing image by transform it from the spatial domain to the frequency domain.(4) The research on wavelet Transform which is application on removing thin cloud of satellite remote sensing image.(5) Propose a method to remove thin cloud of remote sensing image based on image merge and homomorphism filter theory.(6) The research on removing thin cloud in color image.The conclusions of the article listed as following:Many methods and algorithms have already been proposed to remove the thin cloud of the remote sensing image in the world, but the results are not very well. The image merge method is proposed by this article can remove the thin cloud of remote sensing image and restore the information of non-cloud region.