Tree Crown Delineation Algorithm Based on a Convolutional Neural Network

Author(s)

Braga, Jose R. G. ^[1] ; Peripato, Vinicius ^[1] ; Dalagnol, Ricardo ^[1] ; Ferreira, Matheus P. ^[2] ; Tarabalka, Yuliya ^{[3, 4]} ; Aragao, Luiz E. O. C. ^{[5, 1]} ; de Campos Velho, Haroldo E. ^[6] ; Shiguemori, Elcio H. ^[7] ; Wagner, Fabien H. ^{[1, 8]}

Affiliation

^[1] Natl Inst Space Res INPE, Remote Sensing Div, Av Astronautas 1758, BR-12227010 Sao Jose Dos Campos - Brazil

^[2] Mil Inst Engn IME, Cartog Engn Sect, Praca Gen Tiburcio 80, BR-22290270 Rio De Janeiro - Brazil

^[3] Inria Sophia Antipolis, F-06902 Valbonne - France

^[4] Luxcarta Technol, Parc Activite Argile, Lot 119b, F-06370 Mouans Sartoux - France

^[5] Univ Exeter, Coll Life & Environm Sci, Exeter EX4 4RJ, Devon - England

^[6] Natl Inst Space Res INPE, Associated Lab Comp & Appl Math, Av Astronautas 1758, BR-12227010 Sao Jose Dos Campos - Brazil

^[7] Inst Adv Studies IEAv, Dept Aerosp Sci & Technol, Trevo Coronel Aviador Jose Alberto Albano Amarant, BR-12228001 Sao Jose Dos Campos - Brazil

^[8] Fdn Sci Technol & Space Applicat FUNCATE, GeoProc Div, BR-12210131 Sao Jose Dos Campos - Brazil

• Document type: Journal article
• Source: REMOTE SENSING; v. 12, n. 8 APR 2 2020.
• Web of Science Citations: 14

Abstract

Tropical forests concentrate the largest diversity of species on the planet and play a key role in maintaining environmental processes. Due to the importance of those forests, there is growing interest in mapping their components and getting information at an individual tree level to conduct reliable satellite-based forest inventory for biomass and species distribution qualification. Individual tree crown information could be manually gathered from high resolution satellite images; however, to achieve this task at large-scale, an algorithm to identify and delineate each tree crown individually, with high accuracy, is a prerequisite. In this study, we propose the application of a convolutional neural network-Mask R-CNN algorithm-to perform the tree crown detection and delineation. The algorithm uses very high-resolution satellite images from tropical forests. The results obtained are promising-the Recall, Precision, and F1 score values obtained were were 0.81, 0.91, and 0.86, respectively. In the study site, the total of tree crowns delineated was 59,062. These results suggest that this algorithm can be used to assist the planning and conduction of forest inventories. As the algorithm is based on a Deep Learning approach, it can be systematically trained and used for other regions. (AU)

• FAPESP's process: 16/17652-9 - Functional diversity of intact and regenerating Amazon, Atlantic Forest and Cerrado systems using hyperspectral imagery
• Grantee: Fabien Hubert Wagner
• Support Opportunities: Scholarships in Brazil - Young Researchers
• FAPESP's process: 18/06072-7 - Automatic registration of high resolution UAV images with satellite images in tropical forests
• Grantee: Jose Renato Garcia Braga
• Support Opportunities: Scholarships in Brazil - Technical Training Program - Technical Training
• FAPESP's process: 15/22987-7 - Assessment of climate change impacts on the biomass and carbon dynamics in the Amazon
• Grantee: Ricardo Dal'Agnol da Silva
• Support Opportunities: Scholarships in Brazil - Doctorate
• FAPESP's process: 18/15001-6 - ARBOLES: a trait-based understanding of LATAM forest biodiversity and resilience
• Grantee: Luiz Eduardo Oliveira e Cruz de Aragão
• Support Opportunities: Regular Research Grants
• FAPESP's process: 15/50484-0 - Functional diversity of intact and regenerating Amazon, Atlantic Forest, and Cerrado systems using hyperspectral imagery
• Grantee: Fabien Hubert Wagner
• Support Opportunities: Research Grants - Young Investigators Grants