An Experiment on Discrimination of Close Two Lines by High Precision Hough Transform using Round and Merge Method

Abstract

Hough transform is one of the most stabilized methods for feature extraction from digital images. One of the problems of using Hough transform for practical use is that it requires a lot of memory for the θ-ρ parameter plane. If a very high precision θ or ρ is required, the parameter plane grows very large. In this paper, a method for Hough transform computation without using a parameter plane is proposed and the results of the experiment are shown. By this method, high precision computation can be executed using very little memory.