Image motion degradation compensation for high dynamic imaging of space-based vertical orbit scanning

Abstract

Rotating Payload Satellite (RPS) utilizes payload rotation to drive the optical axis for vertical orbit scanning, which enables high-resolution and wide-coverage imaging of ground curved targets. However, the presence of irregular image motion degradation (IMD) in the dynamic imaging drastically degrades the imaging quality. High stability and high precision IMD compensation have become key point for high-resolution imaging of RPS. In this paper, an IMD compensation model is proposed based on velocity vector prediction and multiple disturbance identification. Firstly, time-varying multi-dimensional velocity vectors are analyzed based on the object-to-image mapping relationship. This method is used to predict the rotation angle of the sensor, which can ensure the sensor’s exposure direction always follows the direction of image motion. Then, to enhance accuracy and stability of compensation, the actual angular velocity of sensor rotation is extracted from various disturbance sources through coordinate transformation and provided as feedback. The experiment indicates that the precision and stability of sensor rotation can reach 3.925 × 10^-3 and 8.574 × 10^-4 deg/s. The compensation error is smaller than the threshold of 1/3 pixel. The simulated images of RPS indicate that the deblurring and cumulative deformation correction effects are significant. The image quality is improved by 52.68 % after compensation. It demonstrates that our approach is highly effective and crucial for the practical application of RPS.