The Theory And Software Realization Of Satellite SAR Image Geo-Rectification

Satellite SAR has aptotic earth-circling period, and can cover wide ground area. And with its development, satellite SAR images are widely used in cartography, land use, damage assessment, geognosy, hydrology, ocean, glacier, etc. The geometric rectification of satellite SAR image becomes more and more important in order to avoid geometric distortion and locate the image accurately. This paper designs and realizes a software package for satellite SAR image's geometric rectification.The dissertation introduces the theories of two common methods for geometric rectification presently used—One based on GCP, and another based on DEM. A new algorithm based on simulated imaging point is proposed and tested through combining the two methods. In terms of the requirements of our project, we design the frame and function models of a new geometric-correction software, and discuss the C++ class, data structure and realization mechanism of some core models, such as image access & display model, GCP managing model, rectifying-algorithms model, and project managing model. The principle and flow chart of some key algorithms are also discussed in detail.Through the actual test on RADARSAT images with different rectifying algorithms, the results demonstrate that the software has realized our project's anticipative objects, and it runs well at many aspects, such as convenient operations, project management, enhancing speed and efficiency, etc.