Volume 18, No. 6, 2021
An Enhancing Canny Edge Detection Using SUSAN Filter Based On Deep Learning Algorithm For Improved Image Edge Detection
A. Bhuvaneshwari , Dr. S. Britto Ramesh Kumar
Abstract
In digital images, edges can be detected by applying mathematical techniques in order to identify the locations in which the brightness of an image unexpectedly changes. Edges are generally curved lines segmented into sharp adjustments in picture brightness. This paper proposes a technique for finding sharp and precise edges. Additionally, a deep learning method combined with a Gaussian Canny filter is used to replace the SUSAN filter of detection. This paper aims to evaluate how efficient deep learning is by combining the Canny method and the Smallest Uni value Segment Assimilating filter. Additionally, adding these SUSAN filters will eliminate image noise, and help users detect the corner of pictures for the improved canny algorithm. SUSAN performs better than traditional edge detectors, and produces more accurate edges.
Pages: 4367-4375
Keywords: Canny, SUSAN Filter, Edge Detection, CNN