Fuzzy edge detection based on similarity measure in colour image

Abstract

A novel edge detection technique for color images is presented in this paper. In the proposed method, contemporary fuzzy logic is used to implement a relative pixel similarity value algorithm. The smoothness of each pixel is calculated in four directions by use of weighted similarity rules and is normalized to maximum gray level. In other words, image in three dimensional color spaces is mapped into one dimension. Accordingly, thresholding technique is used to find and highlight the edges associated with the image. The pixels lower than the threshold values are assumed to be edges. Thus with proposed method, edge pixels in a color image are detected simultaneously without any complex calculations such as gradient, Laplace and statistical calculations.