This section has described application of the SUSAN principle to the detection of image edges and lines accurately and quickly. The localization of the features is independent of the mask size used, and noise suppression is good. Connectivity of edges and lines at junctions is good. The SUSAN principle can be viewed as an efficient way of finding features using local information from a pseudo-global viewpoint; it is the image regions which define the features at their edges and boundaries, and the features themselves are not the primary objects of investigation.