A fast billet location algorithm using particle swarm optimization

This paper addresses the problem of real-time location of billet in steel mill. One major difficulty encountered in the billet location control is that there is not a fixed lighting in the kiln, and the light would be changed with the change of temperature. However, the background subtracting method is sensitive to changes of dynamic scene due to lighting and extraneous events which may reduce false detection. This paper proposes an adaptive detection window method based on particle swarm optimization for a video location system with static camera as a visual sensor. In the practical application, the algorithm can adapt the changes of lighting, and obtained good control performance under the industrial environment.