Illumination-invariant line detection with the Gray-scale Hough transform

In this paper, we present an illuminationinvariant gray scale Hough transform that employs a robust edge detection method as a preprocessing. The edge detection is performed based on the analysis of gradient orientations. Consequently, the proposed method can detect lines in an image regardless of changing lighting conditions. Simulation results also show that the method can detect lines successfully even in the image illuminated non-uniformly.