Evaluation of Three Different Operators for Object Highlighting in Medical RGB Images: Canny, Roberts, and LoG in Independent Color Spaces

The research presents a comparative analysis of the selection of objects of interest in color medical cytology images. For these purposes, the individual edge detection operators studied are Canny, Roberts, and LoG. The comparison contrasts the standard method of assessing color photographs with a technique that incorporates supplementary information. Data from individual color channels, derived from the breakdown of a color image in RGB format, serve as supplementary information. The following individual cases for potential examination are emphasized: Image processing with a complete level of reuse of edge detection operators; with a partial level of reuse of edge detection operators. The results are displayed through an array of tables, graphs, and images. Metrics of visual perception of images are utilized for comparative analysis. Compared with traditional edge detection methods, the proposed approach yields better results, which can exceed the estimates of the classical approach by 20%.