Abstract

The problem of edge detection considers two stages: localization and identification, where localization is the search of pixels in an image and identification is the process of deciding if a pixel belongs, or not, to an edge. The Canny edge detector has an effective identification involving the analysis of every pixel that belongs to an image. On the other side, artificial bee colony (ABC) algorithm simulates the foraging behavior of honey bees, doing an efficient search of food sources. In this proposal, ABC algorithm and Canny are integrated to create ABC-ED, an efficient edge detector algorithm, that does not require to analyze all the pixels of an image to detect its edges. The dataset BSDS500 was used for experimentation, and results show that it is not necessary to analyze every pixel in the image to detect the same edges detected when using Canny.