A New Generalized Hough Transformation Algorithm
Gao Shang-yi · Computer Technology and Development · 2008
Generalized hough transform(GHT) is an effective method of detecting irregular objects.While detecting irregular polygon objects,because gradient is the index item in creating reference table,then the number of items of having concrete gradient values is redundant,it results in the large vote redundancy in projecting on parameter space,and affects transform efficiency.But modified GHT algorithm used geometrical feature angle to instead gradient,that caused the number of items to distribute rather evenly,then reduced vote redundancy in parameter space,and improved algorithm's efficiency.Experiments demonstrate that this modified method is effective.