AUTHOR=Connolly Rod M. , Fairclough David V. , Jinks Eric L. , Ditria Ellen M. , Jackson Gary , Lopez-Marcano Sebastian , Olds Andrew D. , Jinks Kristin I. TITLE=Improved Accuracy for Automated Counting of a Fish in Baited Underwater Videos for Stock Assessment JOURNAL=Frontiers in Marine Science VOLUME=8 YEAR=2021 URL=https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2021.658135 DOI=10.3389/fmars.2021.658135 ISSN=2296-7745 ABSTRACT=
The ongoing need to sustainably manage fishery resources can benefit from fishery-independent monitoring of fish stocks. Camera systems, particularly baited remote underwater video system (BRUVS), are a widely used and repeatable method for monitoring relative abundance, required for building stock assessment models. The potential for BRUVS-based monitoring is restricted, however, by the substantial costs of manual data extraction from videos. Computer vision, in particular deep learning (DL) models, are increasingly being used to automatically detect and count fish at low abundances in videos. One of the advantages of BRUVS is that bait attractants help to reliably detect species in relatively short deployments (e.g., 1 h). The high abundances of fish attracted to BRUVS, however, make computer vision more difficult, because fish often obscure other fish. We build upon existing DL methods for identifying and counting a target fisheries species across a wide range of fish abundances. Using BRUVS imagery targeting a recovering fishery species, Australasian snapper (