Fast Line Detection by Hough Transform Using Inter-Image Operations

In image analysis, it is important to detect lines (straight lines) that represent the edges of an object in the image. The most popular technique for line detection is the Hough transform. In this transform, peaks resulting in the accumulator array that are obtained through a voting procedure in the parameter space represent strong evidence that a corresponding line exists in the image. In the voting procedure, large numbers of votes that are far from the peaks are unnecessary and reduce the speed of the transform. In this paper, we propose a method to speed up this process by introducing a procedure in image space that removes lines except for limited directions. This procedure makes it possible to reduce the voting range and to remove unnecessary votes. Experimental results confirm that the proposed method performs effectively.