If you use BarkNet 1.0 or this code in your work, please cite the following article:
https://arxiv.org/abs/1803.00949

@INPROCEEDINGS{8593514, author={M. {Carpentier} and P. {Giguère} and J. {Gaudreault}}, booktitle={2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, title={Tree Species Identification from Bark Images Using Convolutional Neural Networks}, year={2018}, volume={}, number={}, pages={1075-1081}, keywords={feature extraction;forestry;geophysical image processing;image classification;learning (artificial intelligence);neural nets;vegetation mapping;bark images;tree individual number;high-resolution bark images;species recognition;tree diameters;tree bark species classification;standard vision problems;deep learning;forestry related tasks;convolutional neural networks;tree species identification;Vegetation;Forestry;Deep learning;Feature extraction;Training;Cameras;Task analysis}, doi={10.1109/IROS.2018.8593514}, ISSN={2153-0866}, month={Oct},}

Available on Mendeley, in 4 chunks :

• Part 1 of 4: https://data.mendeley.com/datasets/zgr7r2r4nt/1
• Part 2 of 4: https://data.mendeley.com/datasets/xps2rnk8zp/1
• Part 3 of 4: https://data.mendeley.com/datasets/553g3v244g/1
• Part 4 of 4: https://data.mendeley.com/datasets/5sw2c38f8w/7

October 11th, 2019: fixed corrupted BOJ+BOP pictures. https://storage.googleapis.com/barknet-1/BarkNet%201.0001.zip (link expired)

python3 train.py --config PATH_TO_CONFIG_FILE

python3 test.py