Vol. 2, Issue 11, Part D (2016)

Abstract
Edge detection is a fundamental tool in Digital Image Processing. It is widely used for image segmentation in the areas such as image processing, computer vision, and machine vision, particularly for feature detection and feature extraction. Edge detectors are used as pre-processing tools in computer vision that makes image segregation and pattern recognition more comfortable. Many edge detectors are available for pre-processing in computer vision but Canny, Sobel, Prewitt, Roberts and Laplacian of Gaussian (LoG) are among the most applied algorithms. In this paper, each of these algorithms is compared by the manner of comparing the Peak Signal to Noise Ratio (PSNR) and Mean Squared Error (MSE) of the resulting images. Together with PSNR and MSE, time required by each algorithm is also computed to compare these algorithms. The experiment is performed and evaluated on MATLAB. The performance of canny operator is found to be best amongst all the edge detection operators compared in this paper.