In this paper, wood cross-section (CS) segmentation of RGB images is treated. CS segmentation has already been studied for computed tomography images, but few study focuses on RGB images. CS segmentation in rough log ends is an important feature for the both assessment of wood quality and wood traceability. Indeed, it allows to extract other features like pith, eccentricity (distance between the pith and the geometric centre) or annual tree rings which are related to mechanical strength. In image processing, neural networks have been widely used to solve the problem of objects segmentation. In this paper, we propose to compare different state-of-the-art neural networks for CS segmentation task. In particular, we consider U-Net, Mask R-CNN, RefineNet and SegNet. We create an imageset which has been split into 6 subsets . Considered neural networks have been trained on each subset in order to compare their performance on different type of images. Results show different behaviors between neural networks. On the one hand, overall U-Net learns better on small dataset than the others. On the other hand, RefineNet learns well on huge dataset. While SegNet is less efficient and Mask R-CNN does not provide a detailed segmentation. This offers a preliminary result on neural network performances for CS segmentation.

Key messageThe TreeTrace_Douglas database includes images and measurements at several stages of the processing of Douglas fir logs, from sawmill logyard to machine grading and destructive testing of boards, and is suitable for research on quality assessment and traceability. A total of 52 long logs, 156 short logs, 208 wood discs, and 346 boards were analyzed. The image data includes RGB images of log ends and board ends, RGB images and CT slices of strips, and a set of images of the boards (RGB, laser, and X-rays) obtained with an industrial board grading machine. The measurements include wood density, growth ring widths, pith and board location in the logs, heartwood and sapwood areas, mechanical properties of each board obtained by vibratory and static testing, and visual grading of the boards. Dataset is available at https://doi.org/10.15454/YUNEGL and associated metadata are available at https://metadata-afs.nancy.inra.fr/geonetwork/srv/fre/catalog.search#/metadata/d9eef6e4-f195-41f4-b6c2-2ab46adc637e.