Deep BarkID: a portable tree bark identification system by knowledge distillation

Species identification is one of the key steps in the management and conservation planning of many forest ecosystems. We introduce Deep BarkID, a portable tree identification system that detects tree species from bark images. Existing bark identification systems rely heavily on massive computing power access, which may be scarce in many locations. Our approach is deployed as a smartphone application that does not require any connection to a database. Its intended use is in a forest, where internet connection is often unavailable. The tree bark identification is expressed as a bark image classification task, and it is implemented as a convolutional neural network (CNN). This research focuses on developing light-weight CNN models through knowledge distillation. Overall, we achieved 96.12% accuracy for tree species classification tasks for ten common tree species in Indiana, USA. We also captured and prepared thousands of bark images—a dataset that we call Indiana Bark Dataset—and we make it available at https://github.com/wufanyou/DBID .

[1]  Jianping Gou,et al.  Knowledge Distillation: A Survey , 2020, International Journal of Computer Vision.

[2]  Takeshi Saitoh,et al.  CNN-based tree species identification from bark image , 2019, International Conference on Graphic and Image Processing.

[3]  Jeffrey Q. Chambers,et al.  Recognizing Amazonian tree species in the field using bark tissues spectra , 2018, Forest Ecology and Management.

[4]  Shervan Fekri-Ershad,et al.  Bark texture classification using improved local ternary patterns and multilayer neural network , 2020, Expert Syst. Appl..

[5]  Geoffrey E. Hinton,et al.  Distilling the Knowledge in a Neural Network , 2015, ArXiv.

[6]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[7]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[8]  Mark Sandler,et al.  MobileNetV2: Inverted Residuals and Linear Bottlenecks , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[9]  Laure Tougne,et al.  Bark and leaf fusion systems to improve automatic tree species recognition , 2018, Ecol. Informatics.

[10]  Z. Chi,et al.  Bark classification by combining grayscale and binary texture features , 2004, Proceedings of 2004 International Symposium on Intelligent Multimedia, Video and Speech Processing, 2004..

[11]  Itheri Yahiaoui,et al.  Plant identification from bark: A texture description based on Statistical Macro Binary Pattern , 2018, 2018 24th International Conference on Pattern Recognition (ICPR).

[12]  Laure Tougne,et al.  Patch-Based CNN Evaluation for Bark Classification , 2020, ECCV Workshops.

[13]  Jiri Matas,et al.  Fine-grained recognition of plants from images , 2017, Plant Methods.

[14]  Itheri Yahiaoui,et al.  A set of statistical radial binary patterns for tree species identification based on bark images , 2020, Multimedia Tools and Applications.

[15]  R. Sablatnig,et al.  Automated identification of tree species from images of the bark , leaves and needles , 2010 .

[16]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[17]  Thomas Hellström,et al.  Autonomous Forest Vehicles: Historic, envisioned, and state-of-the-art , 2009 .

[18]  Laure Tougne,et al.  Efficient Bark Recognition in the Wild , 2019, VISIGRAPP.

[19]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[20]  Z. Chi,et al.  Plant species recognition based on bark patterns using novel Gabor filter banks , 2003, International Conference on Neural Networks and Signal Processing, 2003. Proceedings of the 2003.

[21]  Julien Marie-Pierre,et al.  Tree species identification on large-scale aerial photographs in a tropical rain forest, French Guiana—application for management and conservation , 2006 .

[22]  Zheru Chi,et al.  Bark texture feature extraction based on statistical texture analysis , 2004, Proceedings of 2004 International Symposium on Intelligent Multimedia, Video and Speech Processing, 2004..

[23]  R. A. Leibler,et al.  On Information and Sufficiency , 1951 .

[24]  Patrick Dallaire,et al.  Tree bark re-identification using a deep-learning feature descriptor , 2020, 2020 17th Conference on Computer and Robot Vision (CRV).

[25]  Alex C. Wiedenhoeft,et al.  Classification of CITES-listed and other neotropical Meliaceae wood images using convolutional neural networks , 2018, Plant Methods.

[26]  Takeshi Saitoh,et al.  Automatic tree species identification from natural bark image , 2020, International Conference on Graphic and Image Processing.

[27]  Geoffrey E. Hinton,et al.  When Does Label Smoothing Help? , 2019, NeurIPS.

[28]  Pinliang Dong,et al.  Individual tree identification using a new cluster-based approach with discrete-return airborne LiDAR data , 2021 .

[29]  Michal Haindl,et al.  Bark recognition using novel rotationally invariant multispectral textural features , 2019, Pattern Recognit. Lett..

[30]  Matic Švab.  Computer-vision-based tree trunk recognition , 2014 .

[31]  Philippe Giguère,et al.  Tree Species Identification from Bark Images Using Convolutional Neural Networks , 2018, 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).