Automatic Sorting of Dwarf Minke Whale Underwater Images
A predictable aggregation of dwarf minke whales (<i>Balaenoptera acutorostrata</i> subspecies) occurs annually in the Australian waters of the northern Great Barrier Reef in June–July, which has been the subject of a long-term photo-identification study. Researchers from the Minke Whale...
Main Authors: | , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-04-01
|
Series: | Information |
Subjects: | |
Online Access: | https://www.mdpi.com/2078-2489/11/4/200 |
_version_ | 1797571053867761664 |
---|---|
author | Dmitry A. Konovalov Natalie Swinhoe Dina B. Efremova R. Alastair Birtles Martha Kusetic Suzanne Hillcoat Matthew I. Curnock Genevieve Williams Marcus Sheaves |
author_facet | Dmitry A. Konovalov Natalie Swinhoe Dina B. Efremova R. Alastair Birtles Martha Kusetic Suzanne Hillcoat Matthew I. Curnock Genevieve Williams Marcus Sheaves |
author_sort | Dmitry A. Konovalov |
collection | DOAJ |
description | A predictable aggregation of dwarf minke whales (<i>Balaenoptera acutorostrata</i> subspecies) occurs annually in the Australian waters of the northern Great Barrier Reef in June–July, which has been the subject of a long-term photo-identification study. Researchers from the Minke Whale Project (MWP) at James Cook University collect large volumes of underwater digital imagery each season (e.g., 1.8TB in 2018), much of which is contributed by citizen scientists. Manual processing and analysis of this quantity of data had become infeasible, and Convolutional Neural Networks (CNNs) offered a potential solution. Our study sought to design and train a CNN that could detect whales from video footage in complex near-surface underwater surroundings and differentiate the whales from people, boats and recreational gear. We modified known classification CNNs to localise whales in video frames and digital still images. The required high classification accuracy was achieved by discovering an effective negative-labelling training technique. This resulted in a less than 1% false-positive classification rate and below 0.1% false-negative rate. The final operation-version CNN-pipeline processed all videos (with the interval of 10 frames) in approximately four days (running on two GPUs) delivering 1.95 million sorted images. |
first_indexed | 2024-03-10T20:34:09Z |
format | Article |
id | doaj.art-a9b6861f5d2b498a985cfe026cea6c03 |
institution | Directory Open Access Journal |
issn | 2078-2489 |
language | English |
last_indexed | 2024-03-10T20:34:09Z |
publishDate | 2020-04-01 |
publisher | MDPI AG |
record_format | Article |
series | Information |
spelling | doaj.art-a9b6861f5d2b498a985cfe026cea6c032023-11-19T21:09:34ZengMDPI AGInformation2078-24892020-04-0111420010.3390/info11040200Automatic Sorting of Dwarf Minke Whale Underwater Images Dmitry A. Konovalov0Natalie Swinhoe1Dina B. Efremova2R. Alastair Birtles3Martha Kusetic4Suzanne Hillcoat5Matthew I. Curnock6Genevieve Williams7Marcus Sheaves8College of Science and Engineering, James Cook University, Townsville, QLD 4181, AustraliaCollege of Science and Engineering, James Cook University, Townsville, QLD 4181, AustraliaFunbox Inc., 119017 Moscow, RussiaCollege of Science and Engineering, James Cook University, Townsville, QLD 4181, AustraliaCollege of Science and Engineering, James Cook University, Townsville, QLD 4181, AustraliaCollege of Science and Engineering, James Cook University, Townsville, QLD 4181, AustraliaCSIRO Land and Water, James Cook University, Townsville, QLD 4811, AustraliaCollege of Science and Engineering, James Cook University, Townsville, QLD 4181, AustraliaCollege of Science and Engineering, James Cook University, Townsville, QLD 4181, AustraliaA predictable aggregation of dwarf minke whales (<i>Balaenoptera acutorostrata</i> subspecies) occurs annually in the Australian waters of the northern Great Barrier Reef in June–July, which has been the subject of a long-term photo-identification study. Researchers from the Minke Whale Project (MWP) at James Cook University collect large volumes of underwater digital imagery each season (e.g., 1.8TB in 2018), much of which is contributed by citizen scientists. Manual processing and analysis of this quantity of data had become infeasible, and Convolutional Neural Networks (CNNs) offered a potential solution. Our study sought to design and train a CNN that could detect whales from video footage in complex near-surface underwater surroundings and differentiate the whales from people, boats and recreational gear. We modified known classification CNNs to localise whales in video frames and digital still images. The required high classification accuracy was achieved by discovering an effective negative-labelling training technique. This resulted in a less than 1% false-positive classification rate and below 0.1% false-negative rate. The final operation-version CNN-pipeline processed all videos (with the interval of 10 frames) in approximately four days (running on two GPUs) delivering 1.95 million sorted images.https://www.mdpi.com/2078-2489/11/4/200computer visiondwarf minke whalesconvolutional neural networksunderwater object classificationimage classificationdeep learning |
spellingShingle | Dmitry A. Konovalov Natalie Swinhoe Dina B. Efremova R. Alastair Birtles Martha Kusetic Suzanne Hillcoat Matthew I. Curnock Genevieve Williams Marcus Sheaves Automatic Sorting of Dwarf Minke Whale Underwater Images Information computer vision dwarf minke whales convolutional neural networks underwater object classification image classification deep learning |
title | Automatic Sorting of Dwarf Minke Whale Underwater Images |
title_full | Automatic Sorting of Dwarf Minke Whale Underwater Images |
title_fullStr | Automatic Sorting of Dwarf Minke Whale Underwater Images |
title_full_unstemmed | Automatic Sorting of Dwarf Minke Whale Underwater Images |
title_short | Automatic Sorting of Dwarf Minke Whale Underwater Images |
title_sort | automatic sorting of dwarf minke whale underwater images |
topic | computer vision dwarf minke whales convolutional neural networks underwater object classification image classification deep learning |
url | https://www.mdpi.com/2078-2489/11/4/200 |
work_keys_str_mv | AT dmitryakonovalov automaticsortingofdwarfminkewhaleunderwaterimages AT natalieswinhoe automaticsortingofdwarfminkewhaleunderwaterimages AT dinabefremova automaticsortingofdwarfminkewhaleunderwaterimages AT ralastairbirtles automaticsortingofdwarfminkewhaleunderwaterimages AT marthakusetic automaticsortingofdwarfminkewhaleunderwaterimages AT suzannehillcoat automaticsortingofdwarfminkewhaleunderwaterimages AT matthewicurnock automaticsortingofdwarfminkewhaleunderwaterimages AT genevievewilliams automaticsortingofdwarfminkewhaleunderwaterimages AT marcussheaves automaticsortingofdwarfminkewhaleunderwaterimages |