Spaceborne Remote Sensing Image Compression Transform Algorithm

Satellite remote sensing is one of the most important ways that we obtain information. And images are the primary information carriers. In recent years, with the development of sensing technology, the resolution of remote sensing image is improving in time, space, spectrum. The amount of data from new type of space remote sensor (like imaging spectrometer and synthetic aperture radar and so on) is increasing. Because of the restrictions of the transmission band and storage space, the image compression technology has been one of the most important technologies in information transmission system currently.The general image compression technology is researched by this dissertation, which analyses the basic requirements of the satellite remote sensing image compression and lucubrate the development of the remote sensing image compression techniques. With analyzing the characteristic of the remote sensing images, lucubrating multi-scale geometric analysis (MGA), specially discussing Contourlet transform which was put forward by M.N.Do and M.Vetterli, Contourlet transform is improved. The new arithmetic has the nature of localization, multi-scale, multi-direction which is inherited from Contourlet transform. It is proved that it is a no redundancy, integer to integer transform and it can show the piecewise smooth contours of the image information very well with less data.This dissertation lucubrates a new method of MGA which is easier and faster. After researching wavelet transform, lifting wavelet, and the detachable two-dimensional wavelet transform, a new method of detachable two-dimensional multi-directional wavelet is put forward. It is faster and easier to realize the image transform and has the better result obviously. Experiments prove that this method can efficiently show images with little data.