An optimum approach for digital image edge detection with artificial neural network

Edge detection is a variety of methods aimed at identifying points in a digital image where image brightness changes sharply or more formally where discontinuities exist. Edge detection is one of the most fundamental tasks in different image processing operations such as pattern recognition, feature extraction and computer vision. It involves various mathematical methods, and the points where the image brightness changes sharply are typically arranged into a series of curved line segments called "edges". In this thesis, the edge of digital images is investigated using artificial neural networks. There are various methods and algorithms for finding the edges of digital images. One of these methods, which has attracted a lot of attention in recent years, is the use of artificial neural networks. Here, the edge finding image obtained from the classical Sobel method is used as a teacher for the neural network. Experimental results show that the use of neural networks provides a significant improvement in image edges obtained by classical methods.