March 12, 2008

Norway scientist introduces automated salmon grading system

A method that identifies fish quality through the fusion of machine vision and pattern recognition methods has been developed by Ekrem Misimi, research scientist of Norway's SINTEF Fisheries and Aquaculture Research.

The new method involves feeding descriptions of the size, colour and shape of salmon to the machine, which will then grade the quality of the fish accordingly through image analysis.

"Machine vision and image analysis will enable us to sort fish into "production", "ordinary" and "superior" classes, while revealing blood in the stomach cavity, with an accuracy of 90 percent. Automation can increase productivity and raise processing rates, while companies can avoid having to establish subsidiaries abroad," said Misimi.

The colour of a salmon is indicative of its quality and the current method at use to distinguish quality is a special ruler and a colour-matching card. Human workers, who are not as fast as machines, grade salmon manually. Also, salmon tends to lose freshness quickly due to pre-death stress and stiffening of flesh before processing.

In addition, a common cause of downgrading is flecks of blood on salmon fillets, gotten through bleeding stomach cavity of the fish.

"The Norwegian fish-processing industry has been slow to introduce modern technology, and the production costs of a kilogram of salmon in this country are an average of NOK 5-10 higher than in countries that compete with us. Exports of processed salmon are also still low, so the industry has a lot to gain by adopting these new methods," said Misimi.