Proportional Voting Method for the Hough Transform
Original Publication Date: 1985-Oct-01
Included in the Prior Art Database: 2005-Feb-19
The present method improves the precision in detection of parameterized shapes in the presence of noise and other perturbation by proportional voting in a Hough space. In the classical Hough transform method votes are cast at discrete points regardless of the n-tuple that describes the object. The present method weights the votes cast in the Hough space according to the proximity of the n-tuples to discrete points in the Hough space. Thus the voting is proportional and the precision of the Hough transform method for detecting shapes in the presence of noise is significantly improved. This increased precision is achieved without the need for a more finely quantized Hough space or the added storage or computation characteristically associated with finer quantization.