Sensing Images Quality Improving And Mending Based On Fusion Technology

With the development of the Spaceflight Remote Sensing technology, the approach and capability to acquire Remote Sensing images have improved significantly. Different kinds of Remote Sensing image have their own character, and are applied to different objects. Multiple Remote Sensing Image Fusion is a Technology to combine information from multiple images of the same scene. When in the adverse condition or one image sensor can not provide enough information we need, the technology can help to get an improved image. So, the problem of the pixel level fusion of different remote sensing images and its applications on improving of the Remote Sensing images quality are discussed in this thesis.Firstly, Multiple Remote Sensing Image fusion and Wavelet Transform are introduced, including theory, character and application; the aims, characteristic and gradation of the remote sensing image fusion are also outlined. Secondly, the methods and rules of image fusion are introduced from two aspects:spatial domain and transform domain. Then two novel fusion rules respectively based on local actively measure match and local weighted average gradient. Finally, in application areas, research contents of this thesis are summarized as follows:1,The local images substitution among different resource RS images are implemented to remove cloud covering effect on RS images. With color balance methods, the color gap after local images substitution is cleared up with good experimental results.2,The thin cloud clearing up methods based on Wavelet Transform are implemented to remove cloud covering effect on single RS images.3,Color Adjustment method of the Sequence Digital Images based on image overlapping are implemented to solve the color gap when the Sequence Digital image mosaics.