Research On Pixel-level Pretreatment Of Remote Sensing Image And Fusion

Due to the rapid development of modern sensor technology, more and more remote sensing data of different times and different airspace are applied to practical applications. Research on distinct sensor imaging (multi-source remote sensing image) data has become an important direction in the field of application of remote sensing data. Among these researches, remote sensing pretreatment and image fusion has been more focused on. Ordinarily, traditional pixel-level remote sensing image fusion algorithm have mainly Linear weighted average, High-pass filter, IHS transform, PCA transform, Wavelet and variety of expanding fusion algorithms. Because Contourlet not only has multiresolution time-frequency analysis features, as a new mathematical theory, but also has a good anisotropy characteristics, so it gets more and more attention in remote sensing image fusion studies. For the purpose of having high-quality images, this paper focuses on pretreatment of remote sensing and Contourlet transformation in the pixel-level image fusion in remote sensing applications.In this paper, research contents and results are as follows:Firstly, remote sensing image fusion is studied, including remote video trans- mission error control and error concealment, geometric deformation of image correction, registration, image enhancement processing.Secondly, various kinds of traditional pixel-level remote sensing image fusion algorithm are summarized. Detailed studies in Contourlet transform theory and its applying to pixel-level remote sensing images fusion are carried out. Advantages and disadvantages of various algorithms are verified and compared. Based on these studies above, wavelet-Contourlet-PCA transform fusion algorithm is proposed in this paper. Simulations prove that this method can make the effect map more clear and contains more information than those of using other fusion algorithm.Finally, a conclusion is made in this paper, and takes an outlook of Contourlet transform in remote sensing image fusion.