ABSTRACT
The Hough transform is designed to find particular shapes at any location and any orientation within an image. This chapter considers the detection of lines and annular rings. Hough transformations for other shapes are possible, but, generally, other approaches are preferred. For example, a Hough transform for a circle is the same as a set of correlations with different sized rings. The correlation method is faster and with the use of composite filtering can become even more efficient. The Hough transform can be applied to any shape, but such an operation may increase the output dimensionality. For example, a circle has a location but also has a radius; thus, there is an extra dimension to accommodate the additional parameter. The Hough transform is designed to find specific geometric shapes such as a line or a curve. The Hough transform can find lines of various thicknesses any orientation, and any length above a minimum.
