Author:
Zheng Xiaoyi,Zan Shikai,Lv Xueying,Zhang Fan,Zhang Liu
Abstract
Image quality and resolution are important factors affecting the application value of remote sensing images. Although increasing the optical aperture of space optical remote sensors (SORSs) improves image resolution, it exacerbates the effects of the space environment on imaging quality. Thus, this study proposes thermal active optical technology (TAO) to enhance image quality while increasing the optical aperture of SORSs by actively correcting in-orbit wavefront aberrations. Replacing traditional wavefront detection and reconstruction with numerical calculation and simulation analysis, more realistic in-orbit SORS wavefront aberrations are obtained. Numerical and finite element analyses demonstrate that nonlinearities in TAO control lead to the failure of traditional wavefront correction algorithms. To address this, we use a neural network algorithm combining CNN and ResNet. Simulation results show that the residual of the systematic wavefront RMS error for SORS reduces to 1/100λ. The static and dynamic modular transfer functions are improved, and the structural similarity index is recovered by over 23%, highlighting the effectiveness of TAO in image quality enhancement. The static and thermal vacuum experiments demonstrate the wide applicability and engineering prospects of TAO.
Funder
National Natural Science Foundation of China
173 Key Projects of Basic Research