Image homogenization based on gray level co-occurrence matrix and region-growing for spaceborne Martian exploration SAR calibration

Researchers use radar echoes to probe the ionosphere, with synthetic aperture radar (SAR) being one of the current methods for Mars ionosphere detection. The accuracy of radar echoes affects the ionosphere detection results, making radiometric calibration of radar echoes necessary. Due to the size of point targets and the harsh Martian environment, deploying artificial calibrators on Mars is challenging. Thus, there is a need for a distributed SAR target radiometric calibration method suitable for the unique Martian landscape. The antenna pattern is a major influencing factor in radiometric calibration. Homogenized targets are essential for spaceborne synthetic aperture radar (SAR) to obtain antenna patterns using flat and uniform areas. Mountainous desert geology is widespread on Mars and can be utilized for radiometric calibration, but these areas are not uniform and require the removal of non-uniform regions. This paper presents a homogenization method that employs the Gray-Level Co-occurrence Matrix (GLCM) and a region-growing algorithm to generate a mask, effectively achieving the uniformization of non-uniform surface targets. Consequently, the coefficient of variation (CV), a metric used to assess data dispersion, exhibited a reduction of 34%.