Improved Canny edge detection algorithm
Lang Wang, Yuan Sun
Abstract
Canny edge detection is a widely used operator in image processing. With the wide application of Canny operator, its problems in noise processing, threshold selection and edge selection are becoming more and more obvious. This paper improves the traditional Canny operator and uses adaptive median filter instead of Gaussian filter to enhance image details. At the same time, the gradient value is calculated by adding the gradient template in the directions of 45° and 135°, so as to improve the edge loss; at the same time, the double threshold algorithm is improved to detect and connect the edges. Through verification, the following conclusions are drawn: compared with the traditional Canny algorithm, the improved algorithm has certain advantages and anti-noise in edge extraction, and has certain practicability.