A sensitive geometric self-calibration method and stability analysis for multiview spaceborne SAR images based on the range-Doppler model

Synthetic aperture radar (SAR) image positioning technology is extensively used in many scientific fields, including land surveying and mapping. Geometric self-calibration can be performed if images are captured in three directions. However, when the number of images is too small, self-calibration of the SAR images based on the range-Doppler (RD) model appears to be inaccurate. Hence, a robust geometric calibration method has an important impact on calibration results. The effectiveness of such a method depends on the validity of the SAR images. This implies that the calibration can algorithmically optimize the images involved in self-calibration such that the calibration results are close to the true unknown parameters. To overcome these inaccuracies in geometric calibration, this study proposes a flexible calibration approach. The determinant and accuracy stabilization factor (ASF) are utilized to filter the images, allowing the evaluation of singular solutions and determining the validity of the SAR images. Experimental results demonstrate that the robustness of the proposed approach. In addition, the slant range equation is suggested as the dominant equation for analyzing image calibration error sources and image capture. It is found that satellite position is the main source of image calibration errors. Therefore, the impact of the satellite position and the associated incidence angle on the calibration is analyzed. The analysis reveals that it is desirable for satellites to capture ipsilateral images with incidence angles greater than 8$°$. This finding justifies the acquisition of SAR images.