Hyperspectral imaging for non-destructive detection of process-related impact damage in whole Atlantic cod (Gadus morhua L.)
This study demonstrates that hyperspectral imaging can effectively distinguish Atlantic cod with process-related impact damage from undamaged controls by detecting elevated muscle hemoglobin levels, although factors like tissue depth and scattering currently limit its ability to accurately quantify damage severity or grade individual fish.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the bustling world of industrial seafood processing, speed and automation are the driving forces, yet they often come at a cost to the product itself. When fish are handled by machines or dropped onto conveyor belts, they can suffer from invisible injuries that ruin their quality long before they reach a consumer's plate. These impacts can cause internal bleeding, break bones, and separate muscle fibers, turning a premium product into something substandard. To catch these defects, the industry has traditionally relied on human inspectors, who look at a small sample of the catch and make subjective judgments. This method is slow, inconsistent, and misses the vast majority of damaged fish. Scientists are now turning to a technology called hyperspectral imaging to solve this problem. Unlike a standard camera that sees only the colors of the rainbow, this specialized camera captures hundreds of specific light wavelengths for every single point in an image. This allows it to see chemical changes inside the fish, such as the presence of blood, without ever cutting the skin or damaging the product. The goal is to create a system that can automatically spot these hidden injuries as fish move down a production line, ensuring that only the highest quality fish are sold.
Researchers in Norway set out to test whether this technology could reliably detect damage in whole Atlantic cod, a popular fish species often farmed in the cold North Atlantic. They created a controlled experiment to simulate the kinds of bumps and drops that happen on fishing boats and in processing plants. Using 120 farmed cod, they established a baseline by scanning each fish before any harm was done. Then, they subjected the fish to a series of impacts using heavy plates dropped from a height of 1.7 meters. The plates weighed 1 kilogram, 2 kilograms, and 4 kilograms, representing different levels of force that a fish might encounter in a real-world setting. A control group of fish was left untouched to serve as a comparison. After the impacts, the fish were scanned again with the hyperspectral camera, which used a specific lighting setup to see through the skin and into the muscle tissue. The researchers then waited six days, during which the fish were stored on ice, before they were filleted and examined by a team of trained human judges. Finally, they took samples of the muscle tissue to measure the exact amount of hemoglobin, the protein that carries oxygen in blood, using chemical analysis.
The results showed that the physical damage was severe and directly related to the weight of the falling plate. The fish hit by the 2-kilogram and 4-kilogram plates suffered significant injuries, including broken spines and large amounts of internal bleeding. In fact, every single fish in the 4-kilogram group had a fractured spine, and 77 percent of those in the 2-kilogram group were similarly injured. The 1-kilogram group was less damaged, with only 10 percent showing spinal fractures, but they still showed signs of muscle separation and skin damage. The human judges confirmed these findings, giving the heavily impacted fish much worse scores for quality, particularly noting the presence of blood spots and separated muscle fibers. The chemical analysis backed this up, revealing that the fish hit by the heavier plates had significantly higher levels of hemoglobin in their muscle tissue compared to the untouched fish. This confirmed that the impacts caused blood vessels to rupture, allowing blood to leak into the meat.
When the researchers looked at the data from the hyperspectral camera, they found a mixed but promising picture. The camera was very good at telling the difference between the damaged fish and the healthy ones. It successfully identified that the impacted fish had higher levels of blood-related signals than the control group. However, the technology struggled to give a precise measurement of exactly how bad the damage was for each individual fish. While the camera could see that blood was present, the readings varied widely even among fish that had been hit with the same force. The researchers determined that this was because the camera was trying to see through the fish's skin and into deep muscle layers, a process complicated by the way light scatters and absorbs in living tissue. The shape of the fish and the thickness of the skin also played a role, making it difficult to get a consistent reading from one fish to the next. Consequently, while the system could not yet act as a perfect ruler to measure the exact amount of blood in every single fish, it proved it could act as a reliable filter to separate damaged batches from healthy ones.
The study concludes that hyperspectral imaging is a powerful tool for the seafood industry, but it is not a magic solution for every problem. It is highly effective at spotting that damage has occurred and can distinguish between a batch of fish that has been roughly handled and one that has been treated gently. This capability could allow factories to sort fish automatically, removing the need for subjective human inspection of every single item. However, the technology is not yet ready to tell a processor exactly how many milligrams of blood are in a specific fish or to grade the severity of damage on a precise scale for individual fish. The challenges of light scattering through skin and the complex geometry of a whole fish mean that the readings will always have some uncertainty. For now, the best use of this technology is to provide a quick, objective check on the overall quality of a catch, ensuring that the fish moving forward in the production line are free from the hidden injuries that so often go unnoticed.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.