Multi‑frequency Jitter Detection and Modeling for Remote Sensing Satellite Platforms Using Non‑collinear TDI CCD

The advancement of imaging resolution has made the impact of multi-frequency composite jitter in satellite platforms on non-collinear time delay and integration （TDI） charge-coupled device （CCD） imaging systems increasingly critical. Moreover， the accuracy of jitter detection is constrained by the limited inter-chip overlap region inherent to non-collinear TDI CCDs. To address these challenges， a multi-frequency jitter detection method is proposed， achieving sub-pixel level error extraction. Furthermore， a multi-frequency jitter fitting approach utilizing a scale-adjustable sliding window is introduced. For composite multi-frequency jitter， spectral analysis decomposes the relative jitter error curve， while the scale-adjustable sliding window enables frequency-division fitting and modeling. Validation experiments using Gaofen-8 （GF-8） remote sensing satellite imagery detected jitter at 0.65， 20， and 100 Hz in the cross-track direction and at 0.5， 100， and 120 Hz in the along-track direction， demonstrating the method’s precision in detecting platform jitter at sub-pixel accuracy （<0.2 pixels） and its efficacy in fitting and modeling for non-collinear TDI CCD imaging systems subject to multi-frequency jitter.