The challenge for automatic detection of nematodes has been getting the fish fillets through the system at the industrial speed of 40 cm/s. This problem has now been solved. The images are analysed automatically in a purpose-designed computer program, which registers the position of each nematode as well distribution in belly, loin and tail.
Detecting nematodes after trimming too
The results in this enhanced version agree with earlier trials. This system can achieve similar or better results than manual inspection of cod fillets on the candling table. After manual inspection and trimming, the inspection machine found 33 % of the remaining nematodes. The number of false positives of nematodes was also significantly reduced. The system seems promising for incorporation into new filleting lines for white fish.
“The fact that the inspection machines detected nematodes that were not detected during trimming indicates output equivalent to that achieved during industrial trimming of fillet,” says Project Manager and Scientist Agnar H. Sivertsen. “This is despite the fact that the fish fillets are only scanned with the fillet side up. Consequently, one way to improve the automatic detection of nematodes may be to scan the fillets from both sides.”
In combination with manual trimming
The system should be incorporated after pre-trimming directly on the production line and can be used to divide the production flow in two – one clean stream going directly to portioning and packing and the other to manual trimming to remove nematodes detected by the machine.
“It is now the equipment manufacturers’ responsibility to further develop and make use of this technology,” says Sivertsen.
The project was funded by FHF - The Fishery and Aquaculture Industry Research Fund.
Ny teknologi for kvalitetskontroll i filetproduksjon
