Nowadays, we have a much better understanding of the connection between what we eat and how we feel. For example, if you have a heavy lunch, you might hit a wall while trying to be productive in the afternoon. Whereas if you swap the carbs for lean protein and vegetables, you are much more likely to power through, finishing everything you need to do with time to spare.
Consumers today are much more knowledgeable when it comes to food. They know about dietary needs, nutritional information, and any allergies or intolerances they might have. Plus, they are always looking for the best produce they can find in terms of quality and ethical practices. They want to search for the best ingredients for their homecooked dishes and know they are supporting sustainable farming.
The food industry can take advantage of consumers seeking the best possible produce by implementing advanced quality control and grading for different markets. However, if there are opportunities at the high end, there are also consequences at the low end.
Producers also need quality and safety controls to guarantee all their products meet a certain standard and, more importantly, are safe to consume. Providing contaminated food leads to foodborne illnesses, opening a business up to legal problems, additional financial costs, and reputational damage.
The food industry must find a way to gauge the quality of its products and guarantee safety while also running efficient operations and not letting costs spiral out of control.
Measuring food quality and safety
Quality and safety can be determined by various attributes related to food products. These include:
- Physical: texture, colour, tenderness, etc.
- Chemical: moisture, fat content, pH, etc.
- Biological: bacterial count, etc.
Typical quality and safety controls will involve human inspection and some form of chemical or biological testing requiring laboratory analysis. These processes are labour-intensive, destructive, and slow. It may take a day or longer to return analysis, increasing the time it takes for food to reach consumers.
The industry needs a faster method for testing the quality and safety of food. One that is non-chemical or biological but ensures accuracy, helping suppliers cater to different sectors and expand their market share while also guaranteeing safety.
Combining imaging and spectroscopy to determine food quality
Chemical and biological analysis is too slow, but techniques that assess food by measuring light show significant promise while potentially providing real-time data.
Traditional cameras can image food, delivering colour and spatial (shape, size, etc.) information. However, they are limited to external characteristics and only offer three data points per pixel (Red, Green, and Blue). With limited spectral data, obtaining the information necessary for accurate quality and safety controls is not possible.
In contrast, spectroscopy offers a way to quantify the chemical composition of food. By measuring light reflected off food samples, it is possible to understand the chemistry within, which is ideal for quality and safety checks. Chemical bonds absorb light at particular wavelengths. Within food, absorption is particularly prevalent within the near-infra-red (NIR) spectrum.
Therefore, by shining a known light source and measuring the reflected spectra (intensity vs wavelength), analysts can learn critical information related to food quality, including physical characteristics such as particle size and chemical attributes like moisture.
However, spectroscopy produces a spectrum, not an image (as the name suggests). Without spatial information, its use is limited to homogeneous samples, such as flour. Heterogeneous foods, such as meats, vary in composition. They require a combination of spectral and spatial information to assess the quality present.
The solution is to use Hyperspectral Imaging (HSI), a technique combining spectroscopy and imaging.
What is hyperspectral imaging (HSI)?
HSI generates a spatial map, or image, of spectral variations. It allows for pixel-by-pixel analysis of spectral information in real time. Hyperspectral sensors measure a range of narrow wavelength bands, making it possible to identify absorption from different chemical bonds and develop calibration models to understand food-related properties.
RGB cameras only measure three bands. They condense spectral information into three data points per pixel. These three data points translate to a colour, meaning the output is a 2D colour map or an image. In contrast, a hyperspectral camera outputs a 3D hypercube with considerably more data points per pixel to effectively produce a spectrum for every pixel in the image.
Each layer of the hypercube shows the image corresponding to a particular wavelength band. Combining all of the layers or bands allows analysts to determine absorption as a function of wavelength and position – combining spatial and spectral data to deliver a range of benefits for assessing food quality and safety.
The benefits of hyperspectral imaging in the food industry
- Speed
Whereas chemical analysis takes time to return a result, HSI is essentially real-time. In the past, HSI techniques combined images at different wavelengths to build the hypercube. Modern systems can capture all the data in a single image producing a video feed containing hyperspectral information. These systems could be set up to scan a conveyor belt or mounted to a vehicle, providing real-time inspection of food products for providers to identify any poor-quality items.
- Non-destructive
HSI is contactless and non-destructive. It simply shines a wideband light source (sunlight works great) on the food samples, measuring the reflected light using a passive sensor. The sensor has no physical contact with the food, and the light does not damage it in any way. The food does not need to be prepped before measurements are taken, and no waste is generated.
- Versatile
You don’t need one camera to measure moisture and another to detect foreign objects. The same hyperspectral system can be applied to a range of applications. Plus, the camera can be used to cover large areas. HSI is finding widespread use in agriculture, monitoring crop health and checking for early signs of any disease.
Food industry hyperspectral imaging use cases
Quality assessment
HSI can assess food quality for a range of foods, including fruits, vegetables, meats, crops, and other perishable goods. HSI food quality use cases include:
- Various attributes related to fruit and vegetables, such as ripeness and freshness, defects, and chemical components (acidity, sugar content, moisture levels, etc.)
- Predicting meat quality in terms of freshness, tenderness, and other attributes
- Evaluating the quality of crops, including predicting the protein content of wheat
- Going beyond surface level measurements to determine the internal chemistry of food products, examples include measuring the glucose levels of potatoes, or the protein and fat content present in meats
- Identifying anomalies including mould, assessing the amount the mould present and where it is located
Contaminant detection
HSI provides reliable food inspection to identify any foreign objects (metals. plastics, glass, etc.) or contaminants present. Food production typically uses time-consuming laboratory-based testing to identify contaminants or other techniques such as metal detectors. However, these slow down the inspection process and only work for metal objects, respectively.
In contrast, HSI offers real-time detection of a wide range of contaminants. Hyperspectral systems can be installed to enable digital sorters that automatically identify and remove foreign objects.
Additionally, HSI can be used to spot defects in food packaging that may introduce contamination, reducing the shelf life of the product.
Shelf-life prediction
Shelf life is a critical metric to help reduce waste and optimise the sale of perishable foods. HSI can better predict the shelf life of various items by monitoring food properties over time. For example, moisture content and distribution are critical to food quality and how it will degrade over time. While it is a sign of freshness, it also promotes microbial activity that can quicken the time it takes for food to spoil.
Pathogen identification
HSI shows promise in the early detection of pathogens, including bacteria and fungi, within food samples. This process helps prevent the outbreaks of foodborne illnesses such as salmonella and e-coli, protecting consumers from dangerous products and protecting businesses from legal liability.
Food adulteration
Food producers often source ingredients from third-party vendors. HSI can be used to ensure these products are safe and as advertised, identifying any fraudulent activity like mixing inferior or unsafe ingredients. HSI verifies the authenticity and safety of food, even identifying additives used around the world that may be illegal or unacceptable in specific jurisdictions.
Sorting and grading
With many consumers looking for the best possible ingredients, HSI can measure the quality of products for sorting and grading. This allows for produce to be divided into different tiers depending on their quality. Retailers can market the highest quality items at higher prices while pricing lower-graded goods accordingly.
HSI makes it possible to automatically image and sort food based on a variety of quality metrics. With this information, operations can implement an automatic sorting system, optimising production and minimising the need for subjective human decisions and the possible errors that they bring.
In conclusion
HSI is becoming an increasingly valuable tool in the food industry, enhancing quality controls and guaranteeing the safety of products across the supply chain. By combining spectroscopy and imaging, hyperspectral systems deliver the data needed for modern food quality and safety use cases. All while being faster, non-destructive, and more versatile than lab-based testing.
Living Optics is developing a robust, and affordable hyperspectral imaging solution. Powered by advanced machine learning data analysis, our technology provides real-time video feeds with hyperspectral data. Delivering a range of new food safety use cases possible in a single device.
Get in touch today and learn more about how Living Optics is redefining the future of hyperspectral
imaging.
