Advanced geosynchronous radiation imager thermal infrared channel-to-channel co-registration algorithm based on cloud information and phase correlation

The channel-to-channel co-registration is an important performance metric for The Advanced Geosynchronous Radiation Imager (AGRI) of Fengyun-4A (FY-4A), Chinese geostationary meteorological satellite, and large co-registration errors can have a direct impact upon the reliability of the imager’s quantitatively-derived products that rely on combinations of multiple thermal infrared (TIR) channels. Affected products include the cloud mask, fog, and sandstorm. The AGRI Imager TIR channel-to-channel co-registration algorithm based on cloud information and phase correlation is proposed, which allows a systematic calculation of the co-registration error between AGRI TIR channel image pairs. Firstly, the TIR images are preprocessed to obtain the cloud information spatial distribution images. Secondly, the co-registration error is calculated by fitting the low-frequency portion of the phase difference spectrum image of preprocessed image pairs. Finally, the error of the relative misalignment image is compensated in the frequency domain to achieve co-registration. The proposed algorithm is an area-based approach that does not rely on a set of fixed control points that may be impacted by clouds’ presence. In fact, the cloud is the effective information in the algorithm and there is enough cloud information in the AGRI image pairs to calculate the co-registration error. The in-orbit TIR images of AGRI are processed. The results show that the proposed algorithm has better robustness and can better achieve co-registration between TIR image pairs with high precision on the sub-pixel level.