A Multirule-Based Relative Radiometric Normalization for Multisensor Satellite Images

Relative radiometric normalization (RRN) is a widely used method for enhancing the radiometric consistency among multitemporal satellite images. Diverse satellite images enhance the information for observing the Earth's surface and bring additional uncertainties in the applications using multisensor images, such as change detection, multitemporal analysis, and image fusion. To address this challenge, we developed a multirule-based RRN method for multisensor satellite images, which involves the identification of spectral- and spatial-invariant pseudo-invariant features (PIFs) and a partial least-squares (PLS) regression-based RRN modeling using neighboring target pixels around PIFs. The proposed RRN method was validated on four datasets and demonstrated excellent effectiveness in identifying high-quality PIFs with spectral- and spatial-invariant properties, estimating precise regression models, and enhancing the radiometric consistency of reference-target image pair. Our method outperformed six RRN methods and effectively processed well-registered medium- and high-resolution images from the same sensor. This letter highlights the potential of our method for generating more comparable bitemporal multisensor images.

Automated summary

Relative radiometric normalization (RRN) is an effective method for enhancing radiometric consistency among multitemporal satellite images. In this study, we propose a multirule-based RRN method that identifies spectral- and spatial-invariant pseudo-invariant features (PIFs) and uses partial least-squares (PLS) regression to model RRN, resulting in improved radiometric consistency between reference-target image pairs. Our method outperforms six other RRN methods and shows potential for generating more comparable bitemporal multisensor images.