ABSTRACT
This chapter deals with the main existing methods for the detection of geometric primitives, considering the introductory case, which is lines. In Chapter 3, it was shown how with the help of appropriate filters the magnitude of the gradient and its orientation at a given pixel of the image can be found. Through the application of these filters, it can be decided (perhaps by applying a certain threshold) whether a pixel belongs to the edge of an object. Considering this obtained binary map, lines will appear, along with some other curve that will geometrically describe the type of object contained in the image. The topic of this chapter is, from a binary image produced, to find possible forms contained in it that could be attributed to previously specified types of objects.
