Edge Detecting and Partitioning by Comparative Gradient

Abstract

In this article, we present an efficient and highly accuracy method for edges and parts of a colorful picture extracting. In this method besides the use of common methods for edging, a new process has been added to this stage and noticing to the differences of color parts' amounts, we extracted the edges. In continuance, we introduce three solutions for extracting the parts of an image, noticing to the gotten edges. The first is finding a spot inside each part and then extracting the girth of each part with the place where the lines that are omitted of that spot and the edges cross each other. The second is based on using the line chasing system and joining the ends of the two line chaser, which start their opposite departure and at last join together. And the third, attempt to extract parts noticing to the juncture of bright spots in the shape and by observing the color.