Multi-frame image restoration method for novel rotating synthetic aperture imaging system

Xiyang Zhi,Shikai Jiang,Lei Zhang,Jianming Hu,Lijian Yu,Xinhui Song,Jinnan Gong

DOI: https://doi.org/10.1016/j.rinp.2021.103991

IF: 4.565

2021-04-01

Results in Physics

Abstract:<p>The novel rotating synthetic aperture (RSA) optical imaging system is an important development direction for future high-resolution optical remote sensing satellites in geostationary orbit. However, owing to the rotating rectangular pupil, the point spread function of the RSA system has an asymmetric spatial distribution, and the images obtained using the primary mirror from different rotation angles have nonuniform blur degradation. Moreover, platform vibration and pupil rotation have coupling effects on the RSA imaging, resulting in further radiometric and geometric quality degradation. To address these problems, the image degradation characteristics are first analyzed according to the imaging mechanism. Then, combined with the theory of mutual information, an image registration method is suggested by introducing the orientation gradient information. From this, a multi-frame image restoration model is proposed based on the directional gradient prior of the RSA system image. From the perspective of interpretation and application, when the aspect ratio is less than 3, the proposed inversion restoration method can achieve a satisfactory processing performance. This work can provide engineering application reference for the future space application of RSA imaging technology.</p>

physics, multidisciplinary,materials science

What problem does this paper attempt to address?

This paper attempts to solve the problems encountered by the Rotating Synthetic Aperture (RSA) optical imaging system in obtaining high - resolution images. Specifically, due to the rotation characteristics of the rectangular primary mirror, the Point Spread Function (PSF) of the RSA system has an asymmetric spatial distribution, resulting in non - uniform blurring degradation in images obtained from different rotation angles. In addition, there is a coupling effect between platform vibration and pupil rotation, which further leads to the degradation of radiation and geometric quality. These problems limit the application of RSA imaging technology in high - resolution optical remote - sensing satellites.

To address these challenges, the author first analyzes the characteristics of image degradation according to the imaging mechanism. Then, combined with the mutual information theory, a multi - frame image restoration method based on the directional gradient prior is proposed. This method realizes the accurate registration of images obtained by the rectangular primary mirror at different rotation angles by introducing directional gradient information, and on this basis, an image restoration model is constructed to fuse high - resolution information in different directions. Experimental results show that when the aspect ratio of the rectangular pupil is less than 3, the proposed restoration method can achieve satisfactory processing results, providing an engineering reference for the future space application of RSA imaging technology.