Space shift variant image degradation model and restoration of image motion blur in space borne TDI imaging system

摘要

Image motion blur caused by jitters of platform and other factors in space borne TDI (Time delay and integration)imaging system normally is space shift variant. This situation is different from common cases where the image degradation caused by motion is modeled as space invariant and can be dealt with by de-convolution techniques. A space shift variant image degradation model of image motion blur in TDI imaging system is presented using piecewise linear interpolation. The parameter of the model is focal image motion which can be measured or detected in several ways in satellite. A new restoration algorithm for space shift variant blurred image is also presented. The proposed algorithm is based on piecewise bilinear interpolation and restores image pixel by pixel. The algorithm restores each pixel of degraded image individually. Therefore it is not involved in big matrix's computation, which considerably reduces the computation of image restoration. Software simulations are conducted to prove the validity of the presented model and the algorithm. Restorations are performed with simulated images, where image noise is considered. The simulation results has shown that the model and the algorithm perform well in restoring images degraded by image motion which is known or can be detected with an accuracy of 0.3 pixel(3σ).