Optigrader awarded the ECVision Prize 2004
MARCH 4--Optigrader, a system using machine vision and learning software to improve the efficiency of sawmills by measuring and optimizing the edging of rough boards, has been awarded the ECVision (European Network of Excellence on Cognitive Vision Systems; www.ecvision.info) Prize 2004 for best application development in cognitive vision systems. Optigrader, developed by the Finnish machine-vision company inX Systems Ltd. (www.inx.fi), is commercialized from the European Commission-funded KNOT project that was a part of the EUTIST Integrated Machine Vision cluster. Satakunta Polytechnic (www.samk.fi/osata) coordinated the project.
Traditionally, board classification is done manually. A person evaluates the board by verifying the piece of wood, knots, or other failures it may have. Then he decides whether to cut the board shorter to have a better quality and use more resources or to maintain the original length, says Risto Laaksonen, Group Leader of UPM Western Sawmill Group, Finland.
Optigrader measures both faces of a board with color linescan technology and the edge shape with a two-dimensional CCD technology. The data are used to optimize the cutting position of the boards in a successive edger saw. Optigrader enables a sawmill to get the best possible value recovery out of its rough boards. About half of the production of a sawmill passes through an edger saw. Therefore the value recovery of an edger saw is vital to the economic result of a sawmill.
When grading sideboards, the Optigrader system can increase the economic yield by 5%. In some special sideboard dimensions and especially of pine, the increase can be up to 10%--15%. The value of sideboards in a large sawmill is around 20 million euros. Thus, the annual economic benefit amounts to one million euros. Across Europe, this represents a potential saving of 500 million euros, which translate either into extra profit or increased competitiveness. At the same time, wastage and the need for raw timber are both reduced.
The prize will be presented at The 8th European Conference on Computer Vision--ECCV 2004, Prague, May 11-14, 2004. ECVision is a research network, which was formed to promote research, education, and application systems engineering in cognitive AI-enabled computer vision in Europe. This is done through focused networking, multidisciplinary peer interaction, targeted identification of priority issues, and widespread promotion of the area's challenges and successes within both the academic and industrial communities.
