Machine Vision Based Lumber Grain Measurement

In this paper, measurement of wood grains is discussed. A number of image processing techniques are first applied to images of all four long surfaces of sawn lumber. This gives us the rates and directions of the visible grains, which enables us to estimate the actual thickness and 3D directions of wood fibers. These are very important indicators about the strength of the subject. In these experiments, we consider boards sawn from the heart side of the trunks of spruce trees. It is shown, how the position of the wood core can be estimated to few mm accuracy by only analyzing the grains seen on the surface. Also, we show how the measurements of fiber direction and density of annual growth correlate with the measured strength of the lumber.

• Type: Conference
• Year: 2007
• Where: MVA
• Authors: Matti Niskanen, Olli Silvén
• Keywords: Computer Vision, Image Processing Techniques, MVA 2007, Sawn Lumber, Wood Grains