Silhouette vanished contour discovery of aerial view images by exploiting pixel divergence

Abstract

An image's edge detection is the process of finding and pinpointing sharp discontinuities in an image. Detecting the edges of an image significantly reduces the quantity of data and removes unnecessary information while keeping the fundamental structural aspects of an image. Edge detection is essential when it comes to image categorization in computer vision and object identification. The primary goal of this research is to investigate several strategies for edge detection and shadow of objects in aerial view images. Machine vision, face detection, medical imaging, and object detection are just a few examples of applications where image segmentation has been widely utilized. Image segmentation is categorizing or identifying sub-patterns in given an image. Many algorithms and strategies for picture segmentation have been presented to improve segmentation issues in each application area. Techniques such as threshold-based and region-based picture segmentation were used in this study. An edge detection method such as Sobel, Prewitt and Roberts and the Canny approach is applied to the benchmark image and compared with the proposed octagonal pixel divergence edge detection (OPDED) algorithm. Results show that the proposed approach is more effective than the other methods, with a quality image with edges.