Imagine you’re driving in your car on a bright, sunny day. You’re wearing the polarized sunglasses you recently spent big bucks on. You’re loving the experience because the glasses reduce annoying glare and help you see more clearly as you drive. Your joy bubble bursts when you glance at the phone mounted on your windshield. You can’t see the navigation app because the polarized sunglasses make the screen appear blacked out.
The above describes a classic case of expectation-reality mismatch. This phenomenon can also easily creep in when evaluating inspection technologies that use hyperspectral imaging technology – like PPO’s Smart Imaging System. While many vendors will make it sound like their solutions excel at detecting all foreign materials, the reality is that a single system can’t detect absolutely everything that might contaminate your product.
There’s plenty of compelling evidence that hyperspectral imaging technology excels at detecting those hard-to-find low-density foreign materials, but there are a few items that are more challenging for the technology to find. Let’s explore those now.
3 Materials that can Evade Hyperspectral Imaging
1. Materials with similar spectral signatures
A spectral signature is a unique pattern of light absorption and reflection across different wavelengths that helps identify specific materials based on how they interact with light.
Every foreign material has a spectral signature and some are quite similar to one another. When similar spectral signatures are detected it can be difficult for PPO’s spectrometer to distinguish between “good” products and unwanted materials, like bone. PPO’s Smart Imaging System has proven adept at finding bone-in products. But calcified cartilage, which has a very similar spectral signature to bone, is more difficult for hyperspectral imaging technology to find.
PPO’s Smart Imaging System has proven adept at finding bone in product. But calcified cartilage, which has a very similar spectral signature to bone, is more difficult for the hyperspectral imaging technology to find.
It’s worth noting that some low-density materials that hyperspectral imaging is typically good at finding, like cardboard, can take on new characteristics during processing that make them harder to detect. For example, small pieces of cardboard marinating in a meat sample can absorb enough protein, fat and water by volume to acquire a similar spectral signature as the meat. This would make the “marinated” cardboard more difficult for hyperspectral imaging to detect.
2. Highly reflective materials
Very shiny materials will reflect all light that strikes them. For example, a droplet of water, a mirror or a brilliant piece of metal can all have similar reflective properties.
This can lead to problems on the processing line since a freshly sheared stainless steel blade might look quite similar to an ice crystal recessed in a fold of a trim piece. In this case, the hyperspectral imaging technology could produce a false negative (where it doesn’t identify the metal that is present) or a false positive (where it looks at a product that’s simply frozen or wet but thinks it’s seeing metal).
In the case of metal specifically, a metal detector is a modality that excels at finding this type of foreign material. This is why many operations employ a multi-hurdle approach to foreign material detection.
3. Very dark materials
Foreign materials that are very dark will absorb all light and won’t reflect information back to PPO’s detectors. For example, a black rubber gasket, the tip of a black glove, or black cardboard may absorb every wavelength in the NIR (Near InfraRed) spectrum, meaning no photons will travel to the spectrometer to be detected. As a result, these foreign materials could slip through detection.
The good news is not all black items absorb every wavelength in the NIR. This means some dark materials will be quite successfully detected by hyperspectral imaging technology.
Understanding ROI Requires a Complete Picture
Yes, there are foreign materials that challenge the dependable detection offered by hyperspectral imaging. But it’s still highly effective technology for finding many of the low-density materials that plague many meat processing plants.
When calculating the ROI of hyperspectral imaging it’s important to remember that the limitations are minimal, and many operations relying on PPO’s Smart Imaging System have repeatedly found very challenging materials hiding in their product, including plastic, wood, rubber, cardboard and more.
To make an informed decision you need an objective view of both the benefits of PPO’s Smart Imaging System and any trade-off you might need to make. Our experts can help. Contact PPO today to have a realistic conversation about how hyperspectral imaging can work in your plant.