The development of an industrial rule-based computer vision control system

UK Hardboard, Britain´s only manufacturer of hardboard, manufactures seven different types of board having various properties. Wood fibres are produced in a refiner. Currently, the operator visually inspects the fibre quality, and alters the refiner settings accordingly. This has been shown to be an inadequate control method. Fibre quality is believed to be one of the most important factors in the determination of end product quality. Hence, a SERC/DTI funded Teaching Company Scheme was set up to establish the relationships between machine settings and final board quality, and to produce an online computer vision controller for the refiner. In the system developed, computer vision techniques are used to produce repeatable, quantitative measurements of wood fibres. These measurements are then fed to a rule base which produces controlling action instructions for the refiner