The difficulty in finding edges consists on in the one hand in recognizing cohesive areas and defining edges only where a new area begins. On the other hand all existing edges should be marked as continuously as possible.
With successful edge detection certain image objects can be better recognized and separated.
The edges are detected using edge detectors which are also called edge filters. There are positive edges with a transition from black to white and negative edges with a transition from white to black.
Amongst the best known edge detectors are the Sobel, the Laplace, and the Canny filter. Using such edge filters can lead to a significantly reduced data amount to be processed.
Learn more on this topic with our Imaging & Vision Handbook! Order it now for FREE!