Sheng Leslie Joevenller

The rising problem of pine forests experiencing various forms of damage, notably in young forests situated in northern Sweden, is a mounting cause for worry. The culprit behind this damage is Scots pine blister rust, Törskate (in Swedish), a fungus that specifically attacks the tree's trunk and branches. The primary goal of this project is to create a single image segmentation algorithm capable of utilizing CT data to differentiate between resinous wood caused by fungus and non-resinous wood. A critical aspect of this endeavor is the ability to distinguish the resin deposition region, which may resemble the heartwood. The successful segmentation of resin wood serves as a fundamental step towards the potential adoption of an industrial sawmill process, ultimately leading to the optimization of sawmill production and a reduction in the waste of forestry-based resources for sustainable Swedish timber industry.