Effects of Vibration Measurement Error on Remote Sensing Image Restoration

Satellite vibrations would lead to image motion blur. Since the vibration isolators cannot fully suppress the influence of vibrations, image restoration methods are usually adopted, and the vibration characteristics of imaging system are usually required as algorithm inputs for better restoration results, making the vibration measurement error strongly connected to the final outcome. If the measurement error surpass a certain range, the restoration may not be implemented successfully. Therefore it is important to test the applicable scope of restoration algorithms and control the vibrations within the range, on the other hand, if the algorithm is robust, then the requirements for both vibration isolator and vibration detector can be lowered and thus less financial cost is needed. In this paper, vibration-induced degradation is first analyzed, based on which the effects of measurement error on image restoration are further analyzed. The vibration-induced degradation is simulated using high resolution satellite images and then the applicable working condition of typical restoration algorithms are tested with simulation experiments accordingly. The research carried out in this paper provides a valuable reference for future satellite design which plan to implement restoration algorithms.