Because of the color space size, a color image contains too much information. Consequently, before analyzing it, its information has to be reduced without a loss of relevant data. In order to reach that, the multiresolution process is well adapted. Indeed, it enables a spatiocolor approach considering color and spatial distribution at the same time. More precisely, the construction of a color pyramid is presented. Finally, its utilisation during a segmentation step is discussed.
Hubert Konik, Alain Tremeau, Bernard Laget, "Multiresolution Color Image Analysis" in Proc. IS&T 3rd Color and Imaging Conf., 1995, pp 82 - 85, https://doi.org/10.2352/CIC.1995.3.1.art00021