Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery
Ecosystem restoration and reforestation often operate at large scales, whereas monitoring practices are usually limited to spatially restricted field measurements that are (i) time- and labour-intensive, and (ii) unable to accurately quantify restoration success over hundreds to thousands of hectare...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
PeerJ Inc.
2022-10-01
|
| Series: | PeerJ |
| Subjects: | |
| Online Access: | https://peerj.com/articles/14219.pdf |
| _version_ | 1827609138393251840 |
|---|---|
| author | Nicholas C. Galuszynski Robbert Duker Alastair J. Potts Teja Kattenborn |
| author_facet | Nicholas C. Galuszynski Robbert Duker Alastair J. Potts Teja Kattenborn |
| author_sort | Nicholas C. Galuszynski |
| collection | DOAJ |
| description | Ecosystem restoration and reforestation often operate at large scales, whereas monitoring practices are usually limited to spatially restricted field measurements that are (i) time- and labour-intensive, and (ii) unable to accurately quantify restoration success over hundreds to thousands of hectares. Recent advances in remote sensing technologies paired with deep learning algorithms provide an unprecedented opportunity for monitoring changes in vegetation cover at spatial and temporal scales. Such data can feed directly into adaptive management practices and provide insights into restoration and regeneration dynamics. Here, we demonstrate that convolutional neural network (CNN) segmentation algorithms can accurately classify the canopy cover of Portulacaria afra Jacq. in imagery acquired using different models of unoccupied aerial vehicles (UAVs) and under variable light intensities. Portulacaria afra is the target species for the restoration of Albany Subtropical Thicket vegetation, endemic to South Africa, where canopy cover is challenging to measure due to the dense, tangled structure of this vegetation. The automated classification strategy presented here is widely transferable to restoration monitoring as its application does not require any knowledge of the CNN model or specialist training, and can be applied to imagery generated by a range of UAV models. This will reduce the sampling effort required to track restoration trajectories in space and time, contributing to more effective management of restoration sites, and promoting collaboration between scientists, practitioners and landowners. |
| first_indexed | 2024-03-09T07:26:57Z |
| format | Article |
| id | doaj.art-2e86d6baae9f47769db32c609d026d7d |
| institution | Directory Open Access Journal |
| issn | 2167-8359 |
| language | English |
| last_indexed | 2024-03-09T07:26:57Z |
| publishDate | 2022-10-01 |
| publisher | PeerJ Inc. |
| record_format | Article |
| series | PeerJ |
| spelling | doaj.art-2e86d6baae9f47769db32c609d026d7d2023-12-03T06:51:23ZengPeerJ Inc.PeerJ2167-83592022-10-0110e1421910.7717/peerj.14219Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imageryNicholas C. Galuszynski0Robbert Duker1Alastair J. Potts2Teja Kattenborn3Department of Botany, Nelson Mandela University, Gqeberha, South AfricaDepartment of Botany, Nelson Mandela University, Gqeberha, South AfricaDepartment of Botany, Nelson Mandela University, Gqeberha, South AfricaRemote Sensing Centre for Earth System Research (RSC4Earth), Universität Leipzig, Leipzig, GermanyEcosystem restoration and reforestation often operate at large scales, whereas monitoring practices are usually limited to spatially restricted field measurements that are (i) time- and labour-intensive, and (ii) unable to accurately quantify restoration success over hundreds to thousands of hectares. Recent advances in remote sensing technologies paired with deep learning algorithms provide an unprecedented opportunity for monitoring changes in vegetation cover at spatial and temporal scales. Such data can feed directly into adaptive management practices and provide insights into restoration and regeneration dynamics. Here, we demonstrate that convolutional neural network (CNN) segmentation algorithms can accurately classify the canopy cover of Portulacaria afra Jacq. in imagery acquired using different models of unoccupied aerial vehicles (UAVs) and under variable light intensities. Portulacaria afra is the target species for the restoration of Albany Subtropical Thicket vegetation, endemic to South Africa, where canopy cover is challenging to measure due to the dense, tangled structure of this vegetation. The automated classification strategy presented here is widely transferable to restoration monitoring as its application does not require any knowledge of the CNN model or specialist training, and can be applied to imagery generated by a range of UAV models. This will reduce the sampling effort required to track restoration trajectories in space and time, contributing to more effective management of restoration sites, and promoting collaboration between scientists, practitioners and landowners.https://peerj.com/articles/14219.pdfMachine learningRestoration ecologyEcosystem monitoringSpekboomAlbany subtropical thicketDrone imagery |
| spellingShingle | Nicholas C. Galuszynski Robbert Duker Alastair J. Potts Teja Kattenborn Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery PeerJ Machine learning Restoration ecology Ecosystem monitoring Spekboom Albany subtropical thicket Drone imagery |
| title | Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery |
| title_full | Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery |
| title_fullStr | Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery |
| title_full_unstemmed | Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery |
| title_short | Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery |
| title_sort | automated mapping of portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery |
| topic | Machine learning Restoration ecology Ecosystem monitoring Spekboom Albany subtropical thicket Drone imagery |
| url | https://peerj.com/articles/14219.pdf |
| work_keys_str_mv | AT nicholascgaluszynski automatedmappingofportulacariaafracanopiesforrestorationmonitoringwithconvolutionalneuralnetworksandheterogeneousunmannedaerialvehicleimagery AT robbertduker automatedmappingofportulacariaafracanopiesforrestorationmonitoringwithconvolutionalneuralnetworksandheterogeneousunmannedaerialvehicleimagery AT alastairjpotts automatedmappingofportulacariaafracanopiesforrestorationmonitoringwithconvolutionalneuralnetworksandheterogeneousunmannedaerialvehicleimagery AT tejakattenborn automatedmappingofportulacariaafracanopiesforrestorationmonitoringwithconvolutionalneuralnetworksandheterogeneousunmannedaerialvehicleimagery |