AI monitors marine health in Darwin Harbour

The Northern Territory has deployed a Microsoft AI solution to help monitor and manage marine health by rapidly analysing underwater video captured around Darwin Harbour.

The Northern Territory boasts healthy populations of species such as sharks and sawfish, with its fish and marine resources supporting customary use, commercial and recreational fishing as well as tourism. The value of goods and services produced by the Territory’s primary industries and fisheries reaches over half a billion dollars each year.

Unable to enter the water directly because of risks posed by saltwater crocodiles and sharks, scientists from the Department of Primary Industry and Resources (DPIR) previously attached underwater cameras to buoys in protected reefs off Darwin. They then spent tens of hours watching footage in order to spot and count fish.

The new open source AI platform, developed in association with Microsoft and available on GitHub, automates this laborious process of counting local fish stocks by progressively learning to identify different varieties of fish. It is able to analyse hours of video in minutes, freeing up scientists for more valuable ecosystem sustainable management work.

“Cloud computing and AI are combining to support scientists to gain a deeper understanding of fish populations,” said Microsoft Australia National Technology Officer Lee Hickin. “Freed from the mundane aspects of counting and identifying fish, scientists can instead take the insights from the AI solution and focus on making informed decisions that have significant environmental and economic impacts.”

Using Microsoft Azure AI services, the first iteration of the system was up and running in a month. The solution was widely deployed within six months, and its identification powers have been progressively enhanced using machine learning ever since. The AI system is now able to identify a fish in a video with 95–99% accuracy.

DPIR fisheries scientist Dr Shane Penny said that two particular species — the golden snapper and black jewfish — have been closely monitored with the technology, as research has proven that these commercially and recreationally important species have been overfished around the greater Darwin area. DPIR Chief Information Officer Rowan Dollar is meanwhile keen to explore the potential regulatory applications of the technology; for example, keeping an eye on the commercial catch.

“We could look into setting up a camera on a trawler that’s out at sea and doing on-the-fly identification of the catch, so we can start measuring by catch,” he said. “We can start being able to identify that in real time, to help better manage those fisheries.”

Further potential applications across the NT are already being considered, including monitoring feral fish in the freshwater systems and cattle movements across the Territory.

The solution also has global conservation potential as it can be trained to spot an array of different animal and fish species, using techniques similar to those used for facial recognition in social media. The fact that it is open source meanwhile creates potential for similar platforms to be deployed in different settings around the world, to support important scientific endeavours that will benefit the Earth’s environment and humanity.

“It’s important for every government jurisdiction, regardless of who you are or where you are, to be using technology to gather and analyse data,” said Dollar. “It’ll help you be more efficient and give better value back to your stakeholders.”

Please follow us and share on Twitter and Facebook. You can also subscribe for FREE to our weekly newsletter and bimonthly magazine.