Identification of Tree Species in Japanese Forests based on Aerial Photography and Deep Learning

Natural forests are complex ecosystems whose tree species distribution and their ecosystem functions are still not well understood. Sustainable management of these forests is of high importance because of their significant role in climate regulation, biodiversity, soil erosion and disaster prevention among many other ecosystem services they provide. In Japan particularly, natural forests are mainly located in steep mountains, hence the use of aerial imagery in combination with computer vision are important modern tools that can be applied to forest research. Thus, this study constitutes a preliminary research in this field, aiming at classifying tree species in Japanese mixed forests using UAV images and deep learning in two different mixed forest types: a black pine (Pinus thunbergii)-black locust (Robinia pseudoacacia) and a larch (Larix kaempferi)-oak (Quercus mongolica) mixed forest. Our results indicate that it is possible to identify black locust trees with 62.6 % True Positives (TP) and 98.1% True Negatives (TN), while lower precision was reached for larch trees (37.4% TP and 97.7% TN).

[1]  Andreas Kamilaris,et al.  Training Deep Learning Models via Synthetic Data: Application in Unmanned Aerial Vehicles , 2019, CAIP Workshops.

[2]  Andrea Berton,et al.  Forestry applications of UAVs in Europe: a review , 2017 .

[3]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[4]  J. Knight.  ‘From timber to tourism: recommoditizing the Japanese forest’. (Reprint of Development and Change article below.) , 2000 .

[5]  Y. Nobori,et al.  Temporal changes in nitrogen acquisition of Japanese black pine (Pinus thunbergii) associated with black locust (Robinia pseudoacacia) , 2014, Journal of Forestry Research.

[6]  C. Oyarzún,et al.  Forests and water: The value of native temperate forests in supplying water for human consumption , 2006 .

[7]  Anna Barbati,et al.  Knowledge gaps about mixed forests: What do European forest managers want to know and what answers can science provide? , 2018 .

[8]  Vladlen Koltun,et al.  Multi-Scale Context Aggregation by Dilated Convolutions , 2015, ICLR.

[9]  S. Ito,et al.  Effects of previous land-use on plant species diversity in semi-natural and plantation forests in a warm-temperate region in southeastern Kyushu, Japan , 2004 .

[10]  L. Tang,et al.  Drone remote sensing for forestry research and practices , 2015, Journal of Forestry Research.

[11]  Neil A. Dodgson,et al.  Proceedings Ninth IEEE International Conference on Computer Vision , 2003, Proceedings Ninth IEEE International Conference on Computer Vision.

[12]  Juha Hyyppä,et al.  Individual Tree Detection and Classification with UAV-Based Photogrammetric Point Clouds and Hyperspectral Imaging , 2017, Remote. Sens..

[13]  Masato Katoh,et al.  Classifying tree species in a northern mixed forest using high-resolution IKONOS data , 2004, Journal of Forest Research.

[14]  Lian Pin Koh,et al.  Small Drones for Community-Based Forest Monitoring: An Assessment of Their Feasibility and Potential in Tropical Areas , 2014 .

[15]  Jana Müllerová,et al.  Black locust--successful invader of a wide range of soil conditions. , 2015, The Science of the total environment.

[16]  L. Bruzzone,et al.  Tree species classification in the Southern Alps based on the fusion of very high geometrical resolution multispectral/hyperspectral images and LiDAR data , 2012 .

[17]  Andreas Kamilaris,et al.  Simulating Training Data for Deep Learning Models , 2017 .

[18]  A. R. Proto,et al.  Forest and UAV: a bibliometric review , 2016 .

[19]  N. R. Spencer,et al.  Biological control of alien, invasive pine trees (Pinus species) in South Africa. , 2000 .

[20]  Sustainable management of planted landscapes: lessons from Japan , 2012, Biodiversity and Conservation.

[21]  Adrien Michez,et al.  Discrimination of Deciduous Tree Species from Time Series of Unmanned Aerial System Imagery , 2015, PloS one.

[22]  Brendan Mackey,et al.  Forest resilience, biodiversity, and climate change: a synthesis of the biodiversity/resilience/stability relationship in forest ecosystems , 2009 .

[23]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[24]  Yoshua Bengio,et al.  The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

[25]  P. Ódor,et al.  Environmental drivers of the forest regeneration in temperate mixed forests , 2019, Forest Ecology and Management.

[26]  T. Masaki,et al.  Disturbance regimes and community structures of a riparian and an adjacent terrace stand in the Kanumazawa Riparian Research Forest, northern Japan , 2002 .

[27]  Matthias Schubert,et al.  Automated Detection of Conifer Seedlings in Drone Imagery Using Convolutional Neural Networks , 2019, Remote. Sens..

[28]  Costas Armenakis,et al.  RESNET-BASED TREE SPECIES CLASSIFICATION USING UAV IMAGES , 2019, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences.

[29]  Iasonas Kokkinos,et al.  DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs , 2016, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[30]  K. Futai,et al.  Inhibition of the regeneration of Japanese black pine (Pinus thunbergii) by black locust (Robinia pseudoacacia) in coastal sand dunes , 2007, Journal of Forest Research.

[31]  C. Oyarzún,et al.  VALUING ECOSYSTEM SERVICES OF CHILEAN TEMPERATE RAINFORESTS , 2007 .

[32]  D. Soto,et al.  Assessment of ecosystem services as an opportunity for the conservation and management of native forests in Chile , 2009 .

[33]  R. Pu,et al.  A comparative analysis of high spatial resolution IKONOS and WorldView-2 imagery for mapping urban tree species , 2012 .

[34]  Jianguo Huang,et al.  Interspecific variation in growth responses to tree size, competition and climate of western Canadian boreal mixed forests. , 2018, The Science of the total environment.

[35]  Ü. Mander,et al.  Carbon exchange in a hemiboreal mixed forest in relation to tree species composition , 2019, Agricultural and Forest Meteorology.

[36]  D. Richardson,et al.  Invasive agroforestry trees: problems and solutions. , 2004 .

[37]  D. Norton,et al.  Crown‐stem dimension relationships in two New Zealand native forests , 2005 .

[38]  S. Jung,et al.  Reproduction of a Robinia pseudoacacia population in a coastal Pinus thunbergii windbreak along the Kujukurihama Coast, Japan , 2009, Journal of Forest Research.