Improved Edge Detection Approach to Tackle Edge Thickness and Better Edge Connectivity

Abstract

Edge detection is substantial in helping us to pre-process any image for various applications from helping us to detect objects to detecting various medical conditions. The paper tackled one major shortcoming with the currently present system which is edge thickness. To improve there is an implementation of multiple thresholds instead of two thresholds generally used by techniques like that in Canny. The selected method solves multiple problems perfecting the handling of errors and more real to truth results. Our aim of refining the method helps us in better edge detection in images with low contrast as well as medical images like MRIs and X-rays.