Beluga whale detection from sliced aerial remote sensing images using object detection pipelines
The emergence of high-resolution remote sensing imagery greatly facilitates the activities related to conservation biology, including whale counting. As manual annotating is laborious and subjected to human-induced bias, it is necessary to introduce automatic approaches for whale detection from the...
Main Authors: | , , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
University of Waterloo (Waterloo, Ontario, Canada)
2023
|
Subjects: | |
Online Access: | https://openjournals.uwaterloo.ca/index.php/vsl/article/view/5383 https://doi.org/10.15353/jcvis.v8i1.5383 |
id |
ftuniwaterlooojs:oai:openjournals.uwaterloo.ca:article/5383 |
---|---|
record_format |
openpolar |
spelling |
ftuniwaterlooojs:oai:openjournals.uwaterloo.ca:article/5383 2024-05-19T07:38:16+00:00 Beluga whale detection from sliced aerial remote sensing images using object detection pipelines Patel, Muhammed Chen, Xinwei Brubacher, Neil Xu, Linlin Clausi, David A 2023-05-10 application/pdf https://openjournals.uwaterloo.ca/index.php/vsl/article/view/5383 https://doi.org/10.15353/jcvis.v8i1.5383 eng eng University of Waterloo (Waterloo, Ontario, Canada) https://openjournals.uwaterloo.ca/index.php/vsl/article/view/5383/5671 https://openjournals.uwaterloo.ca/index.php/vsl/article/view/5383 doi:10.15353/jcvis.v8i1.5383 Copyright (c) 2023 Muhammed Patel, Xinwei Chen, Neil Brubacher, Linlin Xu, David A Clausi Journal of Computational Vision and Imaging Systems; Vol. 8 No. 1 (2022): Special issue: Proceedings of CVIS 2022; 75-77 2562-0444 10.15353/jcvis.v8i1 info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion Peer-reviewed Article 2023 ftuniwaterlooojs https://doi.org/10.15353/jcvis.v8i1.538310.15353/jcvis.v8i1 2024-05-01T23:58:14Z The emergence of high-resolution remote sensing imagery greatly facilitates the activities related to conservation biology, including whale counting. As manual annotating is laborious and subjected to human-induced bias, it is necessary to introduce automatic approaches for whale detection from the large remote sensing dataset based on machine learning-based techniques. In this paper, we implement two deep neural network-based object detection models (i.e., RetinaNet and faster RCNN) to detect the presence of whale in aerial remote sensing images obtained from a survey conducted on Cumberland Sound Bay, Nunavut in 2014. To tackle the difficulties in effective detection caused by the sparse occurrence of whales in the large image, an image-slicing approach is adopted to increase the ratio between the size of whale sample bounding boxes and the input image of the model. Testing results show that compared to downsample on the original image directly, the proposed image slicing approach boost the detection accuracy significantly. Article in Journal/Newspaper Beluga Beluga whale Beluga* Cumberland Sound Nunavut Waterloo Library Journal Publishing Service (University of Waterloo, Canada) |
institution |
Open Polar |
collection |
Waterloo Library Journal Publishing Service (University of Waterloo, Canada) |
op_collection_id |
ftuniwaterlooojs |
language |
English |
description |
The emergence of high-resolution remote sensing imagery greatly facilitates the activities related to conservation biology, including whale counting. As manual annotating is laborious and subjected to human-induced bias, it is necessary to introduce automatic approaches for whale detection from the large remote sensing dataset based on machine learning-based techniques. In this paper, we implement two deep neural network-based object detection models (i.e., RetinaNet and faster RCNN) to detect the presence of whale in aerial remote sensing images obtained from a survey conducted on Cumberland Sound Bay, Nunavut in 2014. To tackle the difficulties in effective detection caused by the sparse occurrence of whales in the large image, an image-slicing approach is adopted to increase the ratio between the size of whale sample bounding boxes and the input image of the model. Testing results show that compared to downsample on the original image directly, the proposed image slicing approach boost the detection accuracy significantly. |
format |
Article in Journal/Newspaper |
author |
Patel, Muhammed Chen, Xinwei Brubacher, Neil Xu, Linlin Clausi, David A |
spellingShingle |
Patel, Muhammed Chen, Xinwei Brubacher, Neil Xu, Linlin Clausi, David A Beluga whale detection from sliced aerial remote sensing images using object detection pipelines |
author_facet |
Patel, Muhammed Chen, Xinwei Brubacher, Neil Xu, Linlin Clausi, David A |
author_sort |
Patel, Muhammed |
title |
Beluga whale detection from sliced aerial remote sensing images using object detection pipelines |
title_short |
Beluga whale detection from sliced aerial remote sensing images using object detection pipelines |
title_full |
Beluga whale detection from sliced aerial remote sensing images using object detection pipelines |
title_fullStr |
Beluga whale detection from sliced aerial remote sensing images using object detection pipelines |
title_full_unstemmed |
Beluga whale detection from sliced aerial remote sensing images using object detection pipelines |
title_sort |
beluga whale detection from sliced aerial remote sensing images using object detection pipelines |
publisher |
University of Waterloo (Waterloo, Ontario, Canada) |
publishDate |
2023 |
url |
https://openjournals.uwaterloo.ca/index.php/vsl/article/view/5383 https://doi.org/10.15353/jcvis.v8i1.5383 |
genre |
Beluga Beluga whale Beluga* Cumberland Sound Nunavut |
genre_facet |
Beluga Beluga whale Beluga* Cumberland Sound Nunavut |
op_source |
Journal of Computational Vision and Imaging Systems; Vol. 8 No. 1 (2022): Special issue: Proceedings of CVIS 2022; 75-77 2562-0444 10.15353/jcvis.v8i1 |
op_relation |
https://openjournals.uwaterloo.ca/index.php/vsl/article/view/5383/5671 https://openjournals.uwaterloo.ca/index.php/vsl/article/view/5383 doi:10.15353/jcvis.v8i1.5383 |
op_rights |
Copyright (c) 2023 Muhammed Patel, Xinwei Chen, Neil Brubacher, Linlin Xu, David A Clausi |
op_doi |
https://doi.org/10.15353/jcvis.v8i1.538310.15353/jcvis.v8i1 |
_version_ |
1799477674250338304 |