Visual Billet Location Control Using Particle Swarm Optimization in Steel Mill

This paper presents a visual billet location control system based on particle swarm optimization in the heating kiln. There is not a fixed lighting, and the environment illumination would change gradually with the change of temperature. This paper proposes a robust detecting method for detecting the moving billet in the heating kiln. The statistical methods are used to classify the sensed image, three representative images are extracted as the background images for background subtraction to describe the illumination change of environment. The illumination change can detect by matching similitude between the sensed image and three background images using absolute difference and PSO. In the practical experiment, the visual billet location control system can get a well control performance in a steel workshop.