Quantification of resinous wood through X-ray CT imaging segmentation method to provide sawing pattern references for sawmills

Resinous wood development in Scots pine trees affected by blister rust significantly impacts timber quality and is a key factor in log sorting and sawing at sawmills. Assessing timber based solely on external symptoms is often inadequate, as internal characteristics are crucial for determining the best sawing patterns. Advances in imaging now allow for detailed analysis of timber before processing. Under current grading systems, resinous wood is seen as a defect, but accurately measuring its internal location and quantity has been difficult due to a lack of quantitative CT imaging data. The increasing presence of resinous wood, intensified by climate change, presents a growing challenge for the industry. Currently, there is no method to quantify and characterize resinous wood using CT imaging. This study introduces a tailored segmentation method based on the sample ray principle and Gaussian Mixture (GM), optimized for industrial use and automation. Applied to 92,151 CT images of Scots pine logs, the method detected and segmented resinous wood with a validation of precision and accuracy metrics. This segmentation reveals the internal distribution of resinous wood and aids in determining its relative length along the log's longitudinal direction. This helps optimize sawing patterns, reduce waste, and enhance sustainability in timber production through non-destructive testing.