Sustainable fisheries with deep video analytics.
Sustainability is both a mandated objective of modern fishery regulators and an expectation of our communities. Observer programs have traditionally been used to monitor the catch, including species that are accidentally caught (bycatch). However, human observers are expensive, subjective, and only sample a small portion of the fishing activity.
Increasingly, in Australia and around the world, on-board cameras are replacing human observers on fishing vessels, resulting in large amounts of data that then are analysed manually in the office by human experts (analysts). This is still expensive and inefficient because even though an entire fishing trip is recorded, only a small sample of it is usually ever examined.
We have developed video analytic tools to decrease time and effort to process on-vessel camera footage.
Technology
The technology is based on current Machine Learning, Artificial Neural Networks, computer vision and image recognition.
Applications
- The technology will reduce labour costs associated with fishery observers, and video-review. In addition, the compliance and monitoring coverage of a fishery can be increased
- The tools can be applied to target fish, as well as incidental bycatch, such as seabirds and marine mammals
- Our technology is able to identify, count and measure fish catch.
Intellectual property
Software and novel algorithms.
The team
Our team consists of world-renowned computer visualisation scientists and fisheries experts with decades of experience with industry, regulatory and environmental agencies. Our technical expertise includes video analytics and machine learning which enables industry and government cost efficiencies and improved sustainability of our marine living resources.