New possibilities with CT scanning in the forest value chain

• Industrial high-resolution X-ray computed tomography (CT) scanners have recently been installed at several sawmills worldwide for the description of roundwood interior features and external log shape. These CT scanners represent a technological advancement for sawmill businesses that open a way to higher volume and value yields and new production planning strategies. This paper will present an indicative study of innovative use of non-destructive CT log data in a Swedish softwood sawmill, linking high-quality information of the wood material along the wood-value chain. Sawn timber was observed throughout the sawmill process line, i.e. from the log yard through the sawmill process until grading after the timber was dried. Before sawing, the CT scanner scanned the logs and calculated knot measurements from the 3D CT log data of simulated value-optimized center yield. A corresponding set of knot measurements were later calculated from the camera-based grading of the dried timber. Only considering knots from the two sets of measurements, the sawn timber was automatically given a quality assessment based on CT data, by camera-based scanning data, and by manual visual grading for reference. Partial least squares regression was used to create prediction models by correlating the two sets of knot measurements with the automatically determined grade from the dry-sorting. The prediction models tested increased the grading consistency between the grading based on CT data of virtual planks and based on camera data of the same planks. Furthermore, a traceability algorithm was tested as a tool to generate large data sets for future studies.

Subject headings

LANTBRUKSVETENSKAPER  -- Lantbruksvetenskap, skogsbruk och fiske -- Trävetenskap (hsv//swe)

AGRICULTURAL SCIENCES  -- Agriculture, Forestry and Fisheries -- Wood Science (hsv//eng)

Keyword

sawn timber grading

computed tomography

partial least squares

Träteknik

Wood Science and Engineering

Publication and Content Type

vet (subject category)

kon (subject category)