CatchMonitor tracks and counts discarded fish from trawler videos
CatchMonitor: a machine learning system for automated fish discard quantification
Computer Vision and Pattern Recognition
Summary
Fishing boats often throw back fish they don’t want to keep, but counting these discarded fish manually is hard and time-consuming. The authors built a computer system called CatchMonitor that watches video from fishing boats and automatically counts the fish being thrown away. They improved the system’s ability to recognize fish types by teaching it with both labeled and unlabeled examples. They also checked how well CatchMonitor matches up with human experts who do the counting by hand.
What this means in practice
- •For fisheries management teams: Automatically quantify discarded fish in fishing trawler videos to support sustainable fishing regulations and compliance monitoring.
- •For aquatic environmental monitoring agencies: Use automated discard data to improve ecosystem impact assessments of fishing activities without manual video review.
Authors
Geoff French, Michal Mackiewicz, Mark Fisher, Helen Holah, Rebecca Lamb
Abstract
We report on the continued development of CatchMonitor, resulting in a prototype computer vision system designed to automatically quantify discarded fish from video footage collected from Remote Electronic Monitoring (REM) systems on fishing trawlers. The analysis of trawler surveillance footage is a challenging problem due to the real-world conditions on board fishing vessels. Building on our prior work we improve the accuracy of species identification through the application of semi-supervised learning. We utilise a simple and robust object tracking approach, upon which we build our prototype discard quantification system. Finally we analyse the variability of manual discard quantification performed by multiple expert human analysts, using it as a benchmark against which we compare the performance of our system.