Gürültüye Karşı Dayanıklı Bir Kenar Tespit Algoritması Geliştirilmesi

Abstract

Edge detection algorithms are to be among the most common tools for digital image processing procedures. The success of these algorithms influences directly the performance of the applications based on image processing in which these algorithms are used. Under random noise conditions, which are the leading factors influencing aforementioned performance mostly, the accuracy ratios of the edge detection algorithms are decreasing. While the noise levels are increasing, the problem of sortation between noise and the edge pattern, which belongs to the object in the image that required to be detected, appears. Within this scope, a noise depended parametric edge detection model developed by analysing the existing edge detection algorithms. Thus, the variable parameter based edge detection algorithm which provides adaptation according to noise levels is introduced.