Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to UAVs
Detection and classification of tree species from remote sensing data were performed using mainly multispectral and hyperspectral images and Light Detection And Ranging (LiDAR) data. Despite the comparatively lower cost and higher spatial resolution, few studies focused on images captured by Red-Gre...
Main Authors: | , , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-08-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/19/16/3595 |
id |
doaj-def4a9400e6b4a15a0b180e0a8dff68b |
---|---|
record_format |
Article |
spelling |
doaj-def4a9400e6b4a15a0b180e0a8dff68b2020-11-24T21:21:03ZengMDPI AGSensors1424-82202019-08-011916359510.3390/s19163595s19163595Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to UAVsAnderson Aparecido dos Santos0José Marcato Junior1Márcio Santos Araújo2David Robledo Di Martini3Everton Castelão Tetila4Henrique Lopes Siqueira5Camila Aoki6Anette Eltner7Edson Takashi Matsubara8Hemerson Pistori9Raul Queiroz Feitosa10Veraldo Liesenberg11Wesley Nunes Gonçalves12Faculty of Computer Science, 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, 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, BrazilDepartment of Computer Engineering, Dom Bosco Catholic University, Campo Grande 79117-900, BrazilFaculty of Engineering, Architecture and Urbanism and Geography, Federal University of Mato Grosso do Sul, Campo Grande 79070-900, BrazilCPAQ, Federal University of Mato Grosso do Sul, Aquidauana 79200-000, BrazilInstitute of Photogrammetry and Remote Sensing, Technische Universität Dresden, 01062 Dresden, GermanyFaculty of Computer Science, Federal University of Mato Grosso do Sul, Campo Grande 79070-900, BrazilFaculty of Computer Science, Federal University of Mato Grosso do Sul, Campo Grande 79070-900, BrazilDepartment of Electrical Engineering, Pontifical Catholic University of Rio de Janeiro, Rio de Janeiro 22451-900, BrazilDepartment of Forest Engineering, Santa Catarina State University, Lages 88520-000, BrazilFaculty of Computer Science, Federal University of Mato Grosso do Sul, Campo Grande 79070-900, BrazilDetection and classification of tree species from remote sensing data were performed using mainly multispectral and hyperspectral images and Light Detection And Ranging (LiDAR) data. Despite the comparatively lower cost and higher spatial resolution, few studies focused on images captured by Red-Green-Blue (RGB) sensors. Besides, the recent years have witnessed an impressive progress of deep learning methods for object detection. Motivated by this scenario, we proposed and evaluated the usage of Convolutional Neural Network (CNN)-based methods combined with Unmanned Aerial Vehicle (UAV) high spatial resolution RGB imagery for the detection of law protected tree species. Three state-of-the-art object detection methods were evaluated: Faster Region-based Convolutional Neural Network (Faster R-CNN), YOLOv3 and RetinaNet. A dataset was built to assess the selected methods, comprising 392 RBG images captured from August 2018 to February 2019, over a forested urban area in midwest Brazil. The target object is an important tree species threatened by extinction known as <i>Dipteryx alata</i> Vogel (Fabaceae). The experimental analysis delivered average precision around 92% with an associated processing times below 30 miliseconds.https://www.mdpi.com/1424-8220/19/16/3595object-detectiondeep learningremote sensing |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Anderson Aparecido dos Santos José Marcato Junior Márcio Santos Araújo David Robledo Di Martini Everton Castelão Tetila Henrique Lopes Siqueira Camila Aoki Anette Eltner Edson Takashi Matsubara Hemerson Pistori Raul Queiroz Feitosa Veraldo Liesenberg Wesley Nunes Gonçalves |
spellingShingle |
Anderson Aparecido dos Santos José Marcato Junior Márcio Santos Araújo David Robledo Di Martini Everton Castelão Tetila Henrique Lopes Siqueira Camila Aoki Anette Eltner Edson Takashi Matsubara Hemerson Pistori Raul Queiroz Feitosa Veraldo Liesenberg Wesley Nunes Gonçalves Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to UAVs Sensors object-detection deep learning remote sensing |
author_facet |
Anderson Aparecido dos Santos José Marcato Junior Márcio Santos Araújo David Robledo Di Martini Everton Castelão Tetila Henrique Lopes Siqueira Camila Aoki Anette Eltner Edson Takashi Matsubara Hemerson Pistori Raul Queiroz Feitosa Veraldo Liesenberg Wesley Nunes Gonçalves |
author_sort |
Anderson Aparecido dos Santos |
title |
Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to UAVs |
title_short |
Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to UAVs |
title_full |
Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to UAVs |
title_fullStr |
Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to UAVs |
title_full_unstemmed |
Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to UAVs |
title_sort |
assessment of cnn-based methods for individual tree detection on images captured by rgb cameras attached to uavs |
publisher |
MDPI AG |
series |
Sensors |
issn |
1424-8220 |
publishDate |
2019-08-01 |
description |
Detection and classification of tree species from remote sensing data were performed using mainly multispectral and hyperspectral images and Light Detection And Ranging (LiDAR) data. Despite the comparatively lower cost and higher spatial resolution, few studies focused on images captured by Red-Green-Blue (RGB) sensors. Besides, the recent years have witnessed an impressive progress of deep learning methods for object detection. Motivated by this scenario, we proposed and evaluated the usage of Convolutional Neural Network (CNN)-based methods combined with Unmanned Aerial Vehicle (UAV) high spatial resolution RGB imagery for the detection of law protected tree species. Three state-of-the-art object detection methods were evaluated: Faster Region-based Convolutional Neural Network (Faster R-CNN), YOLOv3 and RetinaNet. A dataset was built to assess the selected methods, comprising 392 RBG images captured from August 2018 to February 2019, over a forested urban area in midwest Brazil. The target object is an important tree species threatened by extinction known as <i>Dipteryx alata</i> Vogel (Fabaceae). The experimental analysis delivered average precision around 92% with an associated processing times below 30 miliseconds. |
topic |
object-detection deep learning remote sensing |
url |
https://www.mdpi.com/1424-8220/19/16/3595 |
work_keys_str_mv |
AT andersonaparecidodossantos assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT josemarcatojunior assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT marciosantosaraujo assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT davidrobledodimartini assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT evertoncastelaotetila assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT henriquelopessiqueira assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT camilaaoki assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT anetteeltner assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT edsontakashimatsubara assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT hemersonpistori assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT raulqueirozfeitosa assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT veraldoliesenberg assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs AT wesleynunesgoncalves assessmentofcnnbasedmethodsforindividualtreedetectiononimagescapturedbyrgbcamerasattachedtouavs |
_version_ |
1726001441443676160 |