Algorithms Of Motion Estimation Based On Correlation Speed Measurement And IMC For Remote Sensing Images

The image motion generated by the relative displacement between the ground scene and imaging system is the main cause of lower resolution in the imaging system of remote sensing satellite. In order to achieve high-resolution requirements, sophisticated image motion compensation is indispensable. Through the analysis of image motion in the imaging process of remote sensing camera, the two-dimensional measurement method and the software compensation for remote sensing image motion were researched. Considering the real-time requirements, line template matching, curve fitting and surface fitting were utilized to obtain the length and direction of image motion in the one-dimensional and the two-dimensional displacement field. Simulation tests showed that this image motion measurement method is small in computation, high in precision; software compensation was achieved by use of the de-convolution method for one-dimensional motion-blurred images and the projection restoration method for two-dimensional motion-blurred images. Simulation results showed that software image motion compensation method can accurately compensate for the whole pixel level of image motion amount to ensure the resolution of remote sensing images.