project image

Aerial imagery surveys are commonly used in marine mammal research to determine population size, habitat distribution and habitat use. Analysis of aerial photos involves hours of manually identifying individuals present in each image and converting raw counts into useable biological statistics. Our research proposes the use of deep learning algorithms as an assistive technology to increase the efficiency of the marine mammal research workflow. To test the feasibility of this proposal, the existing YOLOv4 convolutional neural network model was trained to detect belugas, kayaks and motorized boats in unmanned aerial vehicle (UAV) imagery. Computer-based object detection achieved an average precision of 61.17% for belugas, 98.58% for boats, and 95.97% for kayaks. We then tested the performance of computer vision tracking of belugas and manned watercraft in UAV videos using the DeepSORT tracking algorithm, achieving a multiple object tracking accuracy (MOTA) ranging from 37% – 88% and multiple object tracking precision (MOTP) between 63% – 86%. Results from this research indicate that deep learning technology can perform at a similar caliber as human annotators in beluga and watercraft detection and tracking, allowing for larger datasets to be processed within a fraction of the time.

Data and Resources

Raw data


Related Publications


Field Value
Dataset Name Churchill Beluga Boat Drone Imagery
Dataset General Type imagery
Dataset Type Dataset
Dataset Level 1.1-Quality Controlled Data
Program Website
Keyword Vocabulary Polar Data Catalogue, Global Change Master Directory
Keyword Vocabulary URL ;
Theme Freshwater
Remote Sensing
Dataset Status Complete
Maintenance and Update Frequency Not planned
Dataset Last Revision Date 2019-08-09
Dataset DOI 10.34992/sgs5-yw58
Metadata Creation Date 2022
Publisher CanWIN
Field Value
Dataset Collection Start Date 2019-07-28
Dataset Collection End Date 2019-08-09
Spatial regions Churchill
Spatial extent West Bound Longitude -94.3257662171
Spatial extent East Bound Longitude -94.0706395086
Spatial extent South Bound Latitude 58.6686784698
Spatial extent North Bound Latitude 58.8195795207
Field Value
Sample Collection
Activity Collection Type Field Measurement
Preferred citation
Analytical Instrument
Analytical Method
Field Value
License Name Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
Licence Type Open
Embargo Date
Licence URL
Terms of Access

CanWIN datasets are licensed individually, however most are licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) Public License. Details for the licence applied can be found using the Licence URL link provided with each dataset. By using data and information provided on this site you accept the terms and conditions of the License. Unless otherwise specified, the license grants the rights to the public to use and share the data and results derived therefrom as long as the proper acknowledgment is given to the data licensor (citation), that any alteration to the data is clearly indicated, and that a link to the original data and the license is made available.

Terms of Use

By accessing this data you agree to CanWIN's Terms of Use.

Field Value
Dataset Authors
Dataset Authors 1
Harasyn, Madison
Type of Name
Centre for Earth Observation Science - University of Manitoba
Dataset Authors 2
Barber, David
Type of Name
Centre for Earth Observation Science - University of Manitoba
Project Data Curator Chan, Wayne
Project Data Curator email
Project Data Curator Affiliation Centre for Earth Observation Science - University of Manitoba
Field Value
Related Resources
Field Value
Publications 1
Publication Name
Detection and tracking of belugas, kayaks and motorized boats in drone video using deep learning
Identifier Code
Identifier Type
Relationship to this dataset
Resource Type
Online Resource
Publication Type