Comparison of RGB Indices used for Vegetation Studies based on Structured Similarity Index (SSIM)

Remote sensing methods are receiving more and more attention during vegetation studies, thanks to the rapid development of drones. The use of indices created using different bands of the electromagnetic spectrum is currently a common practice in agriculture e.g. normalized vegetation index (NDVI), for which, in addition to the red (R), green (G) and blue (B) bands, in different infrared (IR) ranges used bands are used. In addition, there are many indices in the literature that can only be calculated from the red, green, blue (RGB) bands and are used for different purposes. The aim of our work was to objectively compare and group the RGB indices found in the literature (37 pcs) using an objective mathematical method (structured similarity index; SSIM), as a result of which we classified the individual RGB indices into groups that give the same result. To do this, we calculated the 37 RGB indexes on a test image, and then compared the resulting images in pairs using the structural similarity index method. As a result, 28 of the 37 indexes examined could be narrowed down to 7 groups - that is, the indexes belonging to the groups are the same - while the remaining 9 indexes showed no similarity with any other index.