An edge detection algorithm for steel bar in hot rolling process (ICCAS 2014)

In hot rolling process, camber is one of the most significant defects, where camber is longitudinal curvature in the plan view of a steel bar. To obtain a camber value of the steel bar, the center line of the steel bar is necessary, where the center line can be calculated through the average of both edge lines. Therefore, the edge information of the overall shape in the steel bar is important to determine an exact camber value. In this paper, we propose a new edge detection algorithm for obtaining the exact edge information for the steel bar only using the image data. The proposed edge detection algorithm not only uses the mean filter and the gradient but also employs the least square fitting to get the edge information exactly. In this process, we use the width information of the steel bar to decide the edge information. Experimental results demonstrate that the proposed algorithm finds the exact edge information for the steel bar efficiently.