“This is of great interest to players who deliver and receive fresh fish. We are working to make it possible in the future to examine all fish when it arrives at the quay,” says Karsten Heia with Nofima.
Researchers have previously used the technology to look for nematodes, and are now developing monitoring methods for further quality parameters – for example blood content and freshness – without other means than taking a picture of it.
Great analysis opportunity
Normal colour photos are composed of three colour channels with red, green and blue light, but using a specially designed camera from Norsk Elektro Optikk they can take pictures across a huge 216 channels. The camera can also detect visible and infra-red light. Combined, this opens up a spectrum of different analyses – as long as you know what you’re looking for.
“It’s called imaging spectroscopy. The technology was developed to examine fillets, such as finding the blood content in the fish meat and looking for nematodes,” Heia says.
Norsk Elektro Optikk created their first hyperspectral camera in 2003 – on assignment from Nofima.
Customised for industry from the outset
The machinery is adapted for a conveyor belt speed of 40 cm per second. In other words, an average cod per second.
“This is the speed of current filleting facilities, and it has been important to us to have real working conditions from the outset.
With the information provided by spectroscopy, one can save a lot of unnecessary sorting work all the way from the quay to processing.
“If one could discard fish that is unsuitable for fillet production, one could exploit capacity at filleting facilities far more efficiently,” Heia says.
Interest in analysis
The blood pattern in fish has not always been Heia’s focus. In 1996 he completed his Ph.D in physics at the University of Tromsø. As a physician he has worked with tidal water analyses through remote monitoring of the sea – from an atmospheric perspective. The fact that he now has donned the micro lens to investigate one of the ocean’s most important food sources originates from the same enthusiasm for scientific analysis.
“Understanding how things are designed, why things are what they are, that’s what interests me. Using all of the information that can be retrieved using the devices we use, with the right algorithms we can look inside the fish with a new perspective.
Both white and red fish
The research project on cod and salmon was initiated a year ago, with funding from the Regional Research Fund North. The industry is represented by Norway Seafoods on white fish, and Cermaq on salmon. This is an extension of several projects that were funded by the Fishery and Aquaculture Industry Research Fund, where focus was on measuring remaining shelf life, identifying blood and melanin stains in salmon fillet, and identifying nematodes, black membrane, skin remains and blood stains in white fish.
“Going forward, we will take what we have learned on white fish and apply it to red fish,” Heia says.
“It is particularly issues related to melanin stains that are of interest to the salmon industry. Some producers of smoked salmon purchase up to 20 per cent more salmon that required to compensate for fish that during filleting proves to require trimming for blood and melanin stains. If they can avoid this, they can save substantial amounts of money."
Commercial application
In time the Nofima researchers hope the research can find commercial application. Now they are working with equipment supplier Marel, which supplies machinery to the fish processing industry.
The goal is to develop a commercial solution in order for the sorting of white fish fillet, based on images that detect quality faults such as nematodes, blood, black membrane and splitting, to become a part of fillet production.
Assessing the shelf life of fish using spectroscopy is another possibility, but was not a part of this project.
“There is the opportunity, if the players want to take part. During storage, proteins in the blood change. Blood oxidises, and the composition of information we get from the images can provide retailers with information on how long one can expect the fish to stay fresh, with an accuracy of plus/minus one day.”
