期刊名称：International Journal of Computer Science and Network Solutions
印刷版ISSN：2345-3397
出版年度：2014
卷号：2
期号：11
页码：15-23
出版社：International Journal of Computer Science and Network Solutions
摘要：In this paper, using fuzzy rules, a method is presented for edge detection in grayscale images. The proposedmethod is a fuzzy system and its membership function inputs are fed by the filter disc using a smoothedimage. This membership function includes 8 inputs that obtain their values through the surveys ofsmoothed image pixels by a 3 × 3 mask; the system also includes an output that determines the amount ofedge. In this method, for the classification of pixels and adjusting their results, 12 fuzzy inference ruleswere used to determine the final values of the edge pixels and illustrate in the output. The results of thecomparison of multiple images using fuzzy algorithms indicate that the approach gives better results thanthe above methods in dealing with the irrelevant edges and noise. So that the noise in the image is highlydeleted and the accuracy of locating the edge positions is also greatly enhanced.
