MULTIRESOLUTION HOUGH TRANSFORM - AN EFFICIENT METHOD OF DETECTING PATTERNS IN IMAGES

The Hough transform is highly compute bound and demands a large amount of storage. In this correspondence, a novel multiresolution implementation of the Hough transform has been proposed. To reduce the computing time, it exploits the reduced information content of multiresolution images and accumulator arrays and uses a simple peak detection algorithm. Logarithmic parameter-range reduction ensures faster convergence than other methods. The algorithm takes care of discretization errors.