Evaluating Thermal and Color Sensors for Automating Detection of Penguins and Pinnipeds in Images Collected with an Unoccupied Aerial System
Estimating seabird and pinniped abundance is central to wildlife management and ecosystem monitoring in Antarctica. Unoccupied aerial systems (UAS) can collect images to support monitoring, but manual image analysis is often impractical. Automating target detection using deep learning techniques may...
Published in: | Drones |
---|---|
Main Authors: | , , , , |
Format: | Text |
Language: | English |
Published: |
Multidisciplinary Digital Publishing Institute
2022
|
Subjects: | |
Online Access: | https://doi.org/10.3390/drones6090255 |
id |
ftmdpi:oai:mdpi.com:/2504-446X/6/9/255/ |
---|---|
record_format |
openpolar |
spelling |
ftmdpi:oai:mdpi.com:/2504-446X/6/9/255/ 2023-08-20T04:00:54+02:00 Evaluating Thermal and Color Sensors for Automating Detection of Penguins and Pinnipeds in Images Collected with an Unoccupied Aerial System Jefferson T. Hinke Louise M. Giuseffi Victoria R. Hermanson Samuel M. Woodman Douglas J. Krause 2022-09-15 application/pdf https://doi.org/10.3390/drones6090255 EN eng Multidisciplinary Digital Publishing Institute Drones in Ecology https://dx.doi.org/10.3390/drones6090255 https://creativecommons.org/licenses/by/4.0/ Drones; Volume 6; Issue 9; Pages: 255 automated detection Antarctica drone census image analysis Text 2022 ftmdpi https://doi.org/10.3390/drones6090255 2023-08-01T06:28:49Z Estimating seabird and pinniped abundance is central to wildlife management and ecosystem monitoring in Antarctica. Unoccupied aerial systems (UAS) can collect images to support monitoring, but manual image analysis is often impractical. Automating target detection using deep learning techniques may improve data acquisition, but different image sensors may affect target detectability and model performance. We compared the performance of automated detection models based on infrared (IR) or color (RGB) images and tested whether IR images, or training data that included annotations of non-target features, improved model performance. For this assessment, we collected paired IR and RGB images of nesting penguins (Pygoscelis spp.) and aggregations of Antarctic fur seals (Arctocephalus gazella) with a small UAS at Cape Shirreff, Livingston Island (60.79 °W, 62.46 °S). We trained seven independent classification models using the Video and Image Analytics for Marine Environments (VIAME) software and created an open-access R tool, vvipr, to standardize the assessment of VIAME-based model performance. We found that the IR images and the addition of non-target annotations had no clear benefits for model performance given the available data. Nonetheless, the generally high performance of the penguin models provided encouraging results for further improving automated image analysis from UAS surveys. Text Antarc* Antarctic Antarctic Fur Seals Antarctica Arctocephalus gazella Livingston Island MDPI Open Access Publishing Antarctic Cape Shirreff ENVELOPE(-60.800,-60.800,-62.417,-62.417) Livingston Island ENVELOPE(-60.500,-60.500,-62.600,-62.600) Shirreff ENVELOPE(-60.792,-60.792,-62.459,-62.459) Drones 6 9 255 |
institution |
Open Polar |
collection |
MDPI Open Access Publishing |
op_collection_id |
ftmdpi |
language |
English |
topic |
automated detection Antarctica drone census image analysis |
spellingShingle |
automated detection Antarctica drone census image analysis Jefferson T. Hinke Louise M. Giuseffi Victoria R. Hermanson Samuel M. Woodman Douglas J. Krause Evaluating Thermal and Color Sensors for Automating Detection of Penguins and Pinnipeds in Images Collected with an Unoccupied Aerial System |
topic_facet |
automated detection Antarctica drone census image analysis |
description |
Estimating seabird and pinniped abundance is central to wildlife management and ecosystem monitoring in Antarctica. Unoccupied aerial systems (UAS) can collect images to support monitoring, but manual image analysis is often impractical. Automating target detection using deep learning techniques may improve data acquisition, but different image sensors may affect target detectability and model performance. We compared the performance of automated detection models based on infrared (IR) or color (RGB) images and tested whether IR images, or training data that included annotations of non-target features, improved model performance. For this assessment, we collected paired IR and RGB images of nesting penguins (Pygoscelis spp.) and aggregations of Antarctic fur seals (Arctocephalus gazella) with a small UAS at Cape Shirreff, Livingston Island (60.79 °W, 62.46 °S). We trained seven independent classification models using the Video and Image Analytics for Marine Environments (VIAME) software and created an open-access R tool, vvipr, to standardize the assessment of VIAME-based model performance. We found that the IR images and the addition of non-target annotations had no clear benefits for model performance given the available data. Nonetheless, the generally high performance of the penguin models provided encouraging results for further improving automated image analysis from UAS surveys. |
format |
Text |
author |
Jefferson T. Hinke Louise M. Giuseffi Victoria R. Hermanson Samuel M. Woodman Douglas J. Krause |
author_facet |
Jefferson T. Hinke Louise M. Giuseffi Victoria R. Hermanson Samuel M. Woodman Douglas J. Krause |
author_sort |
Jefferson T. Hinke |
title |
Evaluating Thermal and Color Sensors for Automating Detection of Penguins and Pinnipeds in Images Collected with an Unoccupied Aerial System |
title_short |
Evaluating Thermal and Color Sensors for Automating Detection of Penguins and Pinnipeds in Images Collected with an Unoccupied Aerial System |
title_full |
Evaluating Thermal and Color Sensors for Automating Detection of Penguins and Pinnipeds in Images Collected with an Unoccupied Aerial System |
title_fullStr |
Evaluating Thermal and Color Sensors for Automating Detection of Penguins and Pinnipeds in Images Collected with an Unoccupied Aerial System |
title_full_unstemmed |
Evaluating Thermal and Color Sensors for Automating Detection of Penguins and Pinnipeds in Images Collected with an Unoccupied Aerial System |
title_sort |
evaluating thermal and color sensors for automating detection of penguins and pinnipeds in images collected with an unoccupied aerial system |
publisher |
Multidisciplinary Digital Publishing Institute |
publishDate |
2022 |
url |
https://doi.org/10.3390/drones6090255 |
long_lat |
ENVELOPE(-60.800,-60.800,-62.417,-62.417) ENVELOPE(-60.500,-60.500,-62.600,-62.600) ENVELOPE(-60.792,-60.792,-62.459,-62.459) |
geographic |
Antarctic Cape Shirreff Livingston Island Shirreff |
geographic_facet |
Antarctic Cape Shirreff Livingston Island Shirreff |
genre |
Antarc* Antarctic Antarctic Fur Seals Antarctica Arctocephalus gazella Livingston Island |
genre_facet |
Antarc* Antarctic Antarctic Fur Seals Antarctica Arctocephalus gazella Livingston Island |
op_source |
Drones; Volume 6; Issue 9; Pages: 255 |
op_relation |
Drones in Ecology https://dx.doi.org/10.3390/drones6090255 |
op_rights |
https://creativecommons.org/licenses/by/4.0/ |
op_doi |
https://doi.org/10.3390/drones6090255 |
container_title |
Drones |
container_volume |
6 |
container_issue |
9 |
container_start_page |
255 |
_version_ |
1774721171441319936 |