Keyword: «grayscale»

This article discusses the design and development of a color channel auto-weighting module for converting RGB images to grayscale. The relevance of this work stems from the fact that currently standard algorithms have nuances and use fixed formulas and coefficients, making it impossible to adapt the conversion to each specific image. This can result in blurred boundaries of the object of interest and the loss of significant details. The proposed approach is based on a statistical analysis of the region of interest and the background, allowing each R, G, and B channel to be assigned appropriate coefficients based on their contribution to separating the object from the background.