Deep-learning for position error detection of the secondary mirror in space optical remote sensing system

When space optical remote sensing system works in orbit, it is easy to be affected by the external environment such as heat, gravity and platform jitter, which makes the position of components such as secondary mirror be misaligned, resulting in the degradation of image quality. The traditional position misalignment detection technology has the disadvantages of complex device, time-consuming calculation and low accuracy. A deep learning method using convolutional neural network (CNN) is proposed to predict the positional misalignment of the secondary mirror directly from the defocus point spread function (PSF). The simulation results show that the system can be restored to the original design state under a small dynamic range of position error simply and quickly, which is a great significance for space remote sensing system in-orbit alignment.