This document discusses edge linking and boundary detection techniques. It begins by describing how ideal edge detection would yield only boundary pixels, but in reality noise and other factors mean not all edges are detected. Edge detection is followed by linking procedures to assemble edge pixels into meaningful boundaries. Local processing analyzes pixel neighborhoods to link similar pixels based on gradient strength and direction. The Hough transform is introduced as a global processing technique that links points by determining if they lie on a specified shape curve through a parameter space representation.