Coral reefs are among the most biodiverse, yet most threatened, ecosystems on the planet. Climate change, rising temperatures and coastal developments are rapidly changing coral reefs. To effectively manage coral reefs and develop plausible responses to pressures, scientists and managers require robust and accurate ecological information.
Over the past few years, we have seen a rapid intake of novel automated technologies to monitor coral reef ecosystems. From autonomous underwater vehicles, remotely operated vehicles and smart satellites, these technologies are producing vast amounts of information that can help on the management and governance of coral reefs.
Computer vision is one of the new technologies used for coral reef monitoring. Computer vision is advancing data collection because it can process information fast, accurately and be more consistent than humans.
A recent study led by the Global Change Institute demonstrated that using computer vision methods for the automatic analysis of underwater coral reef images is cheaper and 200x faster than humans.
As institutes and monitoring programs around the world start to develop and apply computer vision techniques, it is important to disseminate the current computer vision platforms that can help supercharge the monitoring of these important ecosystems.
Platforms
1) BenthoBox
Tool to automatically process images of benthic (seabed) habitat taken during ecological surveys.
The software uses computer vision algorithms to recognise ‘tagged’ seabed features such as sand, algae, sponges, and corals.
Tool used in the Australian Institute of Marine Science Long Term Monitoring Program.
2) BIIGLE 2.0
Tool developed for annotating benthic fauna in marine images collections with tools customized to increased efficiency and effectiveness in the manual annotation process.
Paper: https://www.frontiersin.org/articles/10.3389/fmars.2017.00083/full
3) CoralNet
Resource for benthic images analysis that deploys computer vision algorithms for fully and semi-automated annotation of benthic images.
It also serves as a data repository and collaboration platform.
Latest paper using CoralNet: https://www.frontiersin.org/articles/10.3389/fmars.2019.00222/full
4) Squidle+
Tool for managing, exploring and annotating images, video and large-scale mosaics.
Code: https://squidle.org/api/help?template=api_help_page.html
Leave a comment if you know of any other computer vision resources for coral reef imagery.
Follow the progress of FishID; an automatic platform for fish species identification and abundance quantification through this blog or @seabassphd.
Sebastian Lopez (Seabass), is a PhD Candidate at the Australian Rivers Institute where he is developing and applying artificial intelligence tools to monitor fish populations in marine ecosystems.