Forest Restoration Monitoring Protocol with a Low-Cost Remotely Piloted Aircraft: Lessons Learned from a Case Study in the Brazilian Atlantic Forest
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
Forest Restoration Monitoring Protocol with a Low-Cost Remotely Piloted Aircraft : Lessons Learned from a Case Study in the Brazilian Atlantic Forest. / Albuquerque, Rafael; Eduardo Ferreira, Manuel; Olsen, Søren Ingvor; Ricardo Caetano Tymus, Julio; Palheta Balieiro, Cintia; Mansur, Hendrik; Jos Ribeiro Moura, Ciro; Vitor Silva Costa, Jo o; Ruiz Castello Branco, Maur cio; Henrique Grohmann, Carlos.
In: Remote Sensing, Vol. 13, No. 2401, 2401, 2021.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Forest Restoration Monitoring Protocol with a Low-Cost Remotely Piloted Aircraft
T2 - Lessons Learned from a Case Study in the Brazilian Atlantic Forest
AU - Albuquerque, Rafael
AU - Eduardo Ferreira, Manuel
AU - Olsen, Søren Ingvor
AU - Ricardo Caetano Tymus, Julio
AU - Palheta Balieiro, Cintia
AU - Mansur, Hendrik
AU - Jos Ribeiro Moura, Ciro
AU - Vitor Silva Costa, Jo o
AU - Ruiz Castello Branco, Maur cio
AU - Henrique Grohmann, Carlos
PY - 2021
Y1 - 2021
N2 - Traditional forest restoration (FR) monitoring methods employ spreadsheets and photostaken at the ground level. Since remotely piloted aircraft (RPA) generate a panoramic high resolutionand georeferenced view of the entire area of interest, this technology has high potential to improvethe traditional FR monitoring methods. This study evaluates how low-cost RPA data may contributeto FR monitoring of the Brazilian Atlantic Forest by the automatic remote measurement of TreeDensity, Tree Height, Vegetation Cover (area covered by trees), and Grass Infestation. The pointcloud data was processed to map the Tree Density, Tree Height, and Vegetation Cover parameters.The orthomosaic was used for a Random Forest classification that considered trees and grasses as asingle land cover class. The Grass Infestation parameter was mapped by the difference between thisland cover class (which considered trees and grasses) and the Vegetation Cover results (obtained bythe point cloud data processing). Tree Density, Vegetation Cover, and Grass Infestation parameterspresented F_scores of 0.92, 0.85, and 0.64, respectively. Tree Height accuracy was indicated by theError Percentage considering the traditional fieldwork and the RPA results. The Error Percentagewas equal to 0.13 and was considered accurate because it estimated a 13% shorter height for treesthat averaged 1.93 m tall. Thus, this study showed that the FR structural parameters were accuratelymeasured by the low-cost RPA, a technology that contributes to FR monitoring. Despite accuratelymeasuring the structural parameters, this study reinforced the challenge of measuring the Biodiversityparameter via remote sensing because the classification of tree species was not possible. After all, theBrazilian Atlantic Forest is a biodiversity hotspot, and thus different species have similar spectralresponses in the visible spectrum and similar geometric forms. Therefore, until improved automaticclassification methods become available for tree species, traditional fieldwork remains necessary fora complete FR monitoring diagnostic.
AB - Traditional forest restoration (FR) monitoring methods employ spreadsheets and photostaken at the ground level. Since remotely piloted aircraft (RPA) generate a panoramic high resolutionand georeferenced view of the entire area of interest, this technology has high potential to improvethe traditional FR monitoring methods. This study evaluates how low-cost RPA data may contributeto FR monitoring of the Brazilian Atlantic Forest by the automatic remote measurement of TreeDensity, Tree Height, Vegetation Cover (area covered by trees), and Grass Infestation. The pointcloud data was processed to map the Tree Density, Tree Height, and Vegetation Cover parameters.The orthomosaic was used for a Random Forest classification that considered trees and grasses as asingle land cover class. The Grass Infestation parameter was mapped by the difference between thisland cover class (which considered trees and grasses) and the Vegetation Cover results (obtained bythe point cloud data processing). Tree Density, Vegetation Cover, and Grass Infestation parameterspresented F_scores of 0.92, 0.85, and 0.64, respectively. Tree Height accuracy was indicated by theError Percentage considering the traditional fieldwork and the RPA results. The Error Percentagewas equal to 0.13 and was considered accurate because it estimated a 13% shorter height for treesthat averaged 1.93 m tall. Thus, this study showed that the FR structural parameters were accuratelymeasured by the low-cost RPA, a technology that contributes to FR monitoring. Despite accuratelymeasuring the structural parameters, this study reinforced the challenge of measuring the Biodiversityparameter via remote sensing because the classification of tree species was not possible. After all, theBrazilian Atlantic Forest is a biodiversity hotspot, and thus different species have similar spectralresponses in the visible spectrum and similar geometric forms. Therefore, until improved automaticclassification methods become available for tree species, traditional fieldwork remains necessary fora complete FR monitoring diagnostic.
U2 - 10.3390/rs13122401
DO - 10.3390/rs13122401
M3 - Journal article
VL - 13
JO - Remote Sensing
JF - Remote Sensing
SN - 2072-4292
IS - 2401
M1 - 2401
ER -
ID: 272412399