Sydney Fish Markets: Quality assessment of fish using image and video

This project aims to assess fish size, species and freshness using computer vision and artificial intelligence. A computer-based digital system, which provides an objective assessment based on ‘seeing’ and ‘smelling’, will be developed and applied to the emerging e-trade at the Sydney Fish Market.

Challenge

The current purchasing process at Sydney Fish Market is a hands on affair in which traders physically inspect the produce before they decide what to buy. There is no objective benchmark for quality and subjective assessments sometimes lead to disagreements and issues around trust, while paperwork inefficiencies can delay sales.

As the Fish Market moves towards an online trading system, they need to bring transparency to the supply chain and provide traders and consumers alike with immediate, verified information about the origin and quality of produce.

 

Solution

The UTS Tech Lab team will design a digital fish provenance and quality tracking system, using snapper as the test species.

An app will be created to allow fishers to record catch photos stamped with the time and location. AI-based image processing and E-Nose technology will be integrated into an online platform and mobile app for use by both traders and retail customers. The technology will identify fish species, estimate size and evaluate freshness.

The project will also deliver a simplified sales process with a shorter time to market and access to catch lists and a quality assurance system.

Duration
4 years

Academic team
Associate Professor Jian Zhang

Lab
Multimedia Data Analytics

Engagement model
Joint-funded project

Funding
Sydney Fish Market
Food Agility CRC

Future applications
E-trade systems for produce markets

Area of expertise
Artificial intelligence
Data science
Robotics and automation