Geometric modeling of attitude jitter for three-line-array imaging satellites.

Attitude jitter causes image motion and degrades geometric accuracy of high-resolution satellite images. This work studies the mechanism of the attitude jitter effect on the imaging geometry of three-line-array push-broom sensors onboard satellites, which is a typical configuration used for topographic mapping. Based on a rigorous physical imaging model, we derived quantitative models of the geometric effect of attitude jitter in the roll, pitch, and yaw angles on the image distortions of the forward, nadir, and backward view sensors, and the accuracy of the derived models is validated through comprehensive experiments and analyses. The experimental results reveal the following. First, the attitude jitter in the roll angle dominates the cross-track image deviation; it does not affect the along-track geometry of the nadir-view sensor but marginally affects the off-nadir-view sensors, and the image distortions share a linear relationship with the image column coordinates. Second, the attitude jitter in the pitch angle dominates the along-track image deviation, and the image distortions in the off-nadir-view images are relatively larger than those in the nadir-view images. The attitude jitter in the pitch angle does not affect the cross-track geometry of the nadir-view sensor but marginally affects the off-nadir-view sensors, and the image distortions share a linear relationship with the image column coordinates. Finally, the attitude jitter in the yaw angle mainly causes the cross-track image deviation in the off-nadir-view sensors, and the along-track image geometries of all the three view sensors are marginally affected to the same extent by the yaw angle variation.