Salmon sea lice counting is an important problem in aquaculture management. NORCE has developed a deep learning-based louse counting system based on a camera solution developed by ScaleAQ AS. The system can automatically detect fish from underwater images with very high precision, and then louse and other object detections (e.g. lice at different stages, fish-eyes, tails) are performed on the detected fish at acceptable speed and accuracy.
Together with NCE Seafood Innovation, NORCE is mapping the new merging technologies aiming for digitalization of the salmon sea lice counting at fish farms.