Relative Radiometric Normalization

Relative radiometric normalization (RRN) aims to harmonize the brightness values of multiple images of the same scene acquired at different times or by different sensors, enabling accurate comparisons and analyses. Current research focuses on developing robust RRN methods that are less sensitive to changes in the scene (e.g., due to vegetation growth or cloud cover), employing techniques like latent change noise modeling and incorporating geometric information to improve accuracy. These advancements are crucial for various applications, including precision agriculture (improving crop monitoring), remote sensing (enhancing change detection and object classification), and 3D point cloud processing (reducing noise from reflections). Improved RRN techniques ultimately lead to more reliable and accurate results in these fields.