Semantic Segmentation of Tree-Canopy in Urban Environment with Pixel-Wise Deep Learning
Urban forests are an important part of any city, given that they provide several environmental benefits, such as improving urban drainage, climate regulation, public health, biodiversity, and others. However, tree detection in cities is challenging, given the irregular shape, size, occlusion, and co...
Main Authors: | , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-08-01
|
Series: | Remote Sensing |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-4292/13/16/3054 |
id |
doaj-b98ddfc590d441a4b9c201f0367eb663 |
---|---|
record_format |
Article |
spelling |
doaj-b98ddfc590d441a4b9c201f0367eb6632021-08-26T14:17:10ZengMDPI AGRemote Sensing2072-42922021-08-01133054305410.3390/rs13163054Semantic Segmentation of Tree-Canopy in Urban Environment with Pixel-Wise Deep LearningJosé Augusto Correa Martins0Keiller Nogueira1Lucas Prado Osco2Felipe David Georges Gomes3Danielle Elis Garcia Furuya4Wesley Nunes Gonçalves5Diego André Sant’Ana6Ana Paula Marques Ramos7Veraldo Liesenberg8Jefersson Alex dos Santos9Paulo Tarso Sanches de Oliveira10José Marcato Junior11Faculty of Engineering, Architecture and Urbanism and Geography, Federal University of Mato Grosso do Sul, Campo Grande 79070-900, BrazilComputing Science and Mathematics Division, University of Stirling, Stirling FK9 4LA, UKFaculty of Engineering and Architecture and Urbanism, University of Western São Paulo, Rodovia Raposo Tavares, km 572, Bairro Limoeiro 19067-175, BrazilEnvironment and Regional Development Program, University of Western São Paulo, Rodovia Raposo Tavares, km 572, Bairro Limoeiro 19067-175, BrazilEnvironment and Regional Development Program, University of Western São Paulo, Rodovia Raposo Tavares, km 572, Bairro Limoeiro 19067-175, BrazilFaculty of Engineering, Architecture and Urbanism and Geography, Federal University of Mato Grosso do Sul, Campo Grande 79070-900, BrazilEnvironmental Science and Sustainability, INOVISÃO Universidade Católica Dom Bosco, Av. Tamandaré, 6000, Campo Grande 79117-900, BrazilEnvironment and Regional Development Program, University of Western São Paulo, Rodovia Raposo Tavares, km 572, Bairro Limoeiro 19067-175, BrazilForest Engineering Department, Santa Catarina State University, Avenida Luiz de Camões 2090, Lages 88520-000, BrazilDepartment of Computer Science, Universidade Federal de Minas Gerais, Belo Horizonte 31270-901, BrazilFaculty of Engineering, Architecture and Urbanism and Geography, Federal University of Mato Grosso do Sul, Campo Grande 79070-900, BrazilFaculty of Engineering, Architecture and Urbanism and Geography, Federal University of Mato Grosso do Sul, Campo Grande 79070-900, BrazilUrban forests are an important part of any city, given that they provide several environmental benefits, such as improving urban drainage, climate regulation, public health, biodiversity, and others. However, tree detection in cities is challenging, given the irregular shape, size, occlusion, and complexity of urban areas. With the advance of environmental technologies, deep learning segmentation mapping methods can map urban forests accurately. We applied a region-based CNN object instance segmentation algorithm for the semantic segmentation of tree canopies in urban environments based on aerial RGB imagery. To the best of our knowledge, no study investigated the performance of deep learning-based methods for segmentation tasks inside the Cerrado biome, specifically for urban tree segmentation. Five state-of-the-art architectures were evaluated, namely: Fully Convolutional Network; U-Net; SegNet; Dynamic Dilated Convolution Network and DeepLabV3+. The experimental analysis showed the effectiveness of these methods reporting results such as pixel accuracy of 96,35%, an average accuracy of 91.25%, F1-score of 91.40%, Kappa of 82.80% and IoU of 73.89%. We also determined the inference time needed per area, and the deep learning methods investigated after the training proved to be suitable to solve this task, providing fast and effective solutions with inference time varying from 0.042 to 0.153 minutes per hectare. We conclude that the semantic segmentation of trees inside urban environments is highly achievable with deep neural networks. This information could be of high importance to decision-making and may contribute to the management of urban systems. It should be also important to mention that the dataset used in this work is available on our website.https://www.mdpi.com/2072-4292/13/16/3054remote sensingimage segmentationsustainabilityconvolutional neural networkurban environment |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
José Augusto Correa Martins Keiller Nogueira Lucas Prado Osco Felipe David Georges Gomes Danielle Elis Garcia Furuya Wesley Nunes Gonçalves Diego André Sant’Ana Ana Paula Marques Ramos Veraldo Liesenberg Jefersson Alex dos Santos Paulo Tarso Sanches de Oliveira José Marcato Junior |
spellingShingle |
José Augusto Correa Martins Keiller Nogueira Lucas Prado Osco Felipe David Georges Gomes Danielle Elis Garcia Furuya Wesley Nunes Gonçalves Diego André Sant’Ana Ana Paula Marques Ramos Veraldo Liesenberg Jefersson Alex dos Santos Paulo Tarso Sanches de Oliveira José Marcato Junior Semantic Segmentation of Tree-Canopy in Urban Environment with Pixel-Wise Deep Learning Remote Sensing remote sensing image segmentation sustainability convolutional neural network urban environment |
author_facet |
José Augusto Correa Martins Keiller Nogueira Lucas Prado Osco Felipe David Georges Gomes Danielle Elis Garcia Furuya Wesley Nunes Gonçalves Diego André Sant’Ana Ana Paula Marques Ramos Veraldo Liesenberg Jefersson Alex dos Santos Paulo Tarso Sanches de Oliveira José Marcato Junior |
author_sort |
José Augusto Correa Martins |
title |
Semantic Segmentation of Tree-Canopy in Urban Environment with Pixel-Wise Deep Learning |
title_short |
Semantic Segmentation of Tree-Canopy in Urban Environment with Pixel-Wise Deep Learning |
title_full |
Semantic Segmentation of Tree-Canopy in Urban Environment with Pixel-Wise Deep Learning |
title_fullStr |
Semantic Segmentation of Tree-Canopy in Urban Environment with Pixel-Wise Deep Learning |
title_full_unstemmed |
Semantic Segmentation of Tree-Canopy in Urban Environment with Pixel-Wise Deep Learning |
title_sort |
semantic segmentation of tree-canopy in urban environment with pixel-wise deep learning |
publisher |
MDPI AG |
series |
Remote Sensing |
issn |
2072-4292 |
publishDate |
2021-08-01 |
description |
Urban forests are an important part of any city, given that they provide several environmental benefits, such as improving urban drainage, climate regulation, public health, biodiversity, and others. However, tree detection in cities is challenging, given the irregular shape, size, occlusion, and complexity of urban areas. With the advance of environmental technologies, deep learning segmentation mapping methods can map urban forests accurately. We applied a region-based CNN object instance segmentation algorithm for the semantic segmentation of tree canopies in urban environments based on aerial RGB imagery. To the best of our knowledge, no study investigated the performance of deep learning-based methods for segmentation tasks inside the Cerrado biome, specifically for urban tree segmentation. Five state-of-the-art architectures were evaluated, namely: Fully Convolutional Network; U-Net; SegNet; Dynamic Dilated Convolution Network and DeepLabV3+. The experimental analysis showed the effectiveness of these methods reporting results such as pixel accuracy of 96,35%, an average accuracy of 91.25%, F1-score of 91.40%, Kappa of 82.80% and IoU of 73.89%. We also determined the inference time needed per area, and the deep learning methods investigated after the training proved to be suitable to solve this task, providing fast and effective solutions with inference time varying from 0.042 to 0.153 minutes per hectare. We conclude that the semantic segmentation of trees inside urban environments is highly achievable with deep neural networks. This information could be of high importance to decision-making and may contribute to the management of urban systems. It should be also important to mention that the dataset used in this work is available on our website. |
topic |
remote sensing image segmentation sustainability convolutional neural network urban environment |
url |
https://www.mdpi.com/2072-4292/13/16/3054 |
work_keys_str_mv |
AT joseaugustocorreamartins semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT keillernogueira semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT lucaspradoosco semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT felipedavidgeorgesgomes semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT danielleelisgarciafuruya semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT wesleynunesgoncalves semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT diegoandresantana semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT anapaulamarquesramos semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT veraldoliesenberg semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT jeferssonalexdossantos semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT paulotarsosanchesdeoliveira semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning AT josemarcatojunior semanticsegmentationoftreecanopyinurbanenvironmentwithpixelwisedeeplearning |
_version_ |
1721190307221995520 |