Impact of six-degree-of-freedom tremor perturbations in remote sensing platforms on image quality: modeling, simulation, and restoration

Platform tremor perturbations constitute a critical factor inducing geometric distortions and quality degradation in remote sensing imagery, significantly compromising the processing of highresolution images and precision-oriented remote sensing applications. This paper proposes a comprehensive methodology to address geometric distortions caused by six-degree-of-freedom (6- DOF) platform perturbations. The framework spans the complete workflow from modeling to image restoration The proposed framework comprises three principal components: First, we establish a mathematical model characterizing 6-DOF perturbations and analytically derive their impact on object-space-to-image-plane projection. Subsequently, we develop a simulation environment generating both push-broom scanner imagery and reference frame camera images under controlled perturbation conditions. Finally, we implement a restoration pipeline using sparse optical flow estimation for perturbation parameter inversion and targeted image rectification. Experimental results demonstrate robust restoration capabilities for 6-DOF-induced distortions, presenting a practical solution for UAV and satellite-based high-resolution remote sensing systems. Furthermore, the proposed simulation platform enables generation of curated datasets to support future machine learning applications in image restoration.