An Edge Detection Method Based on Pixels Categorization and Complementary Transformation

In this paper, we present a new and promising edge detection approach based on pixels categorization and complementary transformation (PCCT) to improve the classical edge methods such as Prewitt, Sobel, LoG and Canny. We divide a given image on a set of regions where, some of their pixels will undergo a complementary transformation following some specific conditions. To measure the performance of the approach, we evaluate it qualitatively and quantitatively with some famous ones according to Pratt’s Figure Of Merit (PFOM) and Structural Similarity Index Measure (SSIM) metrics. Experiments conducted highlight the enhanced capabilities of the newly developed method.