Automated 3D tree-ring detection and measurement from X-ray computed tomography

Tree ring analysis is essential to reveal the environmental information encoded in wood structure. It provides quantitative data on anatomical structure which can be used, for example, measure impact of fluctuating environment tree growth, support global vegetation models and dendrochronological archaeological wooden artefacts. Currently, several imaging-based methods tree-ring detection feature estimation exist. However, despite advances computer vision edge recognition algorithms, tree-rings mostly limited two-dimensional (2D) datasets performed manually some cases. This paper describes a new approach estimate three-dimensional (3D) rings their width automatically from X-ray computed tomography data. relies modified Canny algorithm, capable detecting fully connected edges throughout image stack. Our results show that this performs well six species having conifer, ring-porous diffuse-porous boundary structures. In our study, denoising proved critical step achieve accurate results. A major advantage procedure it requires very little no user interaction rendering reproducible measurements. As also 3D representations edges, may used future inspection features.