June 29, 2023

New computer-vision system aims to revolutionise shrimp farming practices

The AI-based MonitorShrimp project, led by the Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research (AWI), in collaboration with shrimp farming companies Oceanloop and Sander Holding, is a computer vision system that aims to detect and monitor growth, population size, mortality and stress in farmed shrimp, The Fish Site reported.

This technology enables real-time and highly accurate counting and length measurement of shrimp, even under challenging conditions such as high stock densities and cloudy water, with up to 95% accuracy.

The system not only provides aquaculture companies with crucial insights into the performance of their shrimp stock, but also promotes ethical farming practices. The technology was specifically designed to meet the sustainability and animal welfare objectives of Europe's land-based shrimp farms.

The initial prototype underwent testing at Oceanloop's research and development farm in Kiel, Germany, in 2022. A modern smartphone positioned above the water's surface automatically captured images of the shrimp at one-minute intervals, transmitting the live data to a local server. Advanced computer-vision algorithms were then utilised to count the number of shrimp in each image and measure the length of individual shrimp. The resulting information was relayed to fish-farming software, enabling the optimisation of growth and feeding models based on real-time data.

The system successfully identified visual indicators of stress in shrimp, which Dr Bert Wecker, Co-CEO of Oceanloop, said that this function serves as both an early warning system and a means to validate their stock density model.

Dr Stephan Ende, project coordinator at AWI, said that mortality rates, ranging from 13% to 26% in salmon and potentially up to 50% in shrimp, exceed those of any other livestock, so online monitoring empowers aquaculture companies to swiftly respond to signs of stress among their stock.

Dr Fabian Riedel, co-founder and Co-CEO of Oceanloop, said that the ability to monitor shrimp growth and health with real-time live data not only reduces risks in aquaculture but also provides complete transparency regarding biomass, enabling optimisation of the value chain from farming to processing to sale.

Tomasz Kowalczyk, founder and CEO of NeuroSYS, an expert in AI, machine learning, and digitalisation that contributed to algorithm development for the project, said that their  company is ready to embrace the role of artificial intelligence and deep learning in revolutionising the shrimp-farming sector.

-      The Fish Farmer