Counting and classifying anchovies and other small pelagic fish is a manual, time-consuming task that is prone to error. Fish count per kilogram is a key parameter for determining the price and commercial destination of the catch, but measuring it accurately requires time and labour, slowing down operations at fish markets and processing plants.
In addition, the lack of objective, systematically recorded data makes batch traceability more difficult and limits the ability of fleets and fishers’ associations to optimise the process and demonstrate the value of the product.
AZTI, together with Inser Robótica — a company specialising in robotics integration and end-of-line automation — has developed an automatic anchovy and small pelagic fish counter based on imaging techniques and artificial intelligence. The system can detect and count individual fish and automatically calculate the fish count per kilogram quickly and accurately.
The system is supplied as a turnkey project, with all equipment fully integrated, installed and operational: a modular belt conveyor feeding system, a computer vision enclosure with all its components, a weighing scale and a PC running the artificial intelligence software. All information is stored in a historical database of batches by vessel and species.
On average, the system can count around 60,000 anchovies per day, with a maximum capacity of 108,000, equivalent to counting a box containing 350 to 400 anchovies every two minutes at most. This limit is determined by the feeding rate of the catch rather than by the capacity of the computer vision system itself.
Sectors: Fisheries and aquaculture sector, Technology 4.0
Research lines: Efficient, sustainable fisheries and aquaculture
Research sublines: Digitalisation and artificial intelligence