Dataset: Automation of tree-ring detection and measurements using deep learning

Datasets used to train and evaluate neural network-based implementation that automates detection and measurement of tree-ring boundaries from coniferous species. SplitTrainValDatasets files contain hand-annotated squared images of <em>Picea abies</em> core samples with over 8000 ring boundaries. RealWorldEvaluation.zip contains real-world core samples of 4 conifer species (<em>Picea abies</em>, <em>Abies alba</em>, <em>Pinus sylvestris</em> and <em>Larix decidua</em>) used to evaluate our application and tables with the results of these evaluations. Full code of the application is available at https://github.com/Gregor-Mendel-Institute/TRG-ImageProcessing/tree/master. More details about the datasets and the application can be found in our publication: Poláček, M., Arizpe, A., Hüther, P., Weidlich, L., Steindl, S., &amp; Swarts, K. (2022). Automation of tree-ring detection and measurements using deep learning (p. 2022.01.10.475709). bioRxiv. https://doi.org/10.1101/2022.01.10.475709.