Abstract
The major problem of image edge detection is finding the edge pixels which have the similar intensity with background. Finding an appropriate thresholding technique to result an efficient edge is difficult. Most existing edge detection algorithms are dependent to the method of thresholding. This is the central idea of this paper in which we define edge occurrences for the intensity (monochrome) images using fuzzy sets to diminish the above mentioned problems.
Edge characteristic functions are proposed for detecting edge pixels within a desired block of an (intensity and binary) image based on quadruple child windowing and binary edge patterns. We called these functions QCW-ECF. Fuzzy theory has been also applied to extend the QCW-ECF algorithm to the FQCW-ECF algorithm for edge detection within a block of a fuzzy intensity image. The (F)QCW-ECF algorithm results in a degree of edginess concerning the middle pixel of the processing block.
Edge detection algorithm, FQCW-ECF, based on fuzzy logic with tunable membership functions would be an effective solution for the above mentioned problems. Cyclic coordinate algorithms are adapted to train the FQCW-ECF algorithm so as to minimize a desire performance index via tuning the membership functions of the fuzzy logic system.
Get full access to this article
View all access options for this article.