SPECTRAL IMAGING IN MACHINE VISION SOLUTIONS
Spectral imaging is a rapidly growing inspection method for industries benefiting from machine vision solutions. It provides valuable information by identifying materials based on their chemical properties, not available with other camera technologies. Objects with similar looks but with different chemical composition can be reliably identified and sorted with a hyperspectral camera. This makes spectral imaging the most powerful inspection method in many applications with high commercial value.
Several process industries like food, pharma and recycling have already proven the benefits and value of the spectral imaging technology. Modern industry reveals continuously new use cases to achieve a sustainable society by introducing more efficient processes and higher quality products to the market.
With the performance-to-cost ratio that now meets the industrial requirements, hyperspectral imaging has become a quickly growing part of the machine vision solutions market among other camera technologies such as x-ray, RGB and 3D, as an integrated technology to a conveyor and robotic systems for in-line inspection in realtime.
As a line scan technology, SPECIM spectral cameras suites ideally to any new or existing machine vision system. The camera speed can be adjusted and synchronized with the speed of the production line. The processed data output, like sorting result, can be integrated into existing machine vision systems analogously to other line scan techniques. The entire material stream is chemically inspected with no interference with the material itself. Thus, no waste is caused by sampling and inspection covers 100% of the stream in realtime.
SPECIM is leading the industrial spectral camera market with our FX series cameras, which are designed specifically for meeting the requirements in industrial applications. They provide the flexibility, performance and cost-efficiency required by OEM’s, machine builders and machine vision integrators while allowing efficient scalability of a machine vision solution based on SPECIM technology.
INDUSTRIAL IN-LINE USE CASES
Spectral camera technology measures and analyses the spectrum of light reflected from or transmitted through the material. In the visible region, a very high separation of colors can be achieved. When going beyond the visible region, we see that chemically different materials have unique spectral signatures.
The same cameras can be used for very different purposes. For example, a camera operating in the near-infrared 900-1700 nm region can be used in the recycling industry for identifying different plastics or in the food industry for determining the sugar level in potatoes. An application specific processing model is created for each use case, looking for certain characteristics of plastics instead of potatoes, and vice versa.
SPECIM product portfolio, widest available on the market, includes cameras to all relevant wavelength ranges required in industrial applications.
Many industries benefit from hyperspectral imaging:
Spectral imaging provides sorting machine builders and system integrators a technology which obtains chemical information across the full product stream. Most widely used technologies in such food inspection environment are metal detection and x-ray to provide safety of the produced food. Unlike the traditional color, metal, or x-ray cameras, spectral cameras can reliably identify product stream materials based on biological and chemical content which enable the food industry to optimize product consistency, improve yield, reduce waste and deliver quality.
Meat fish and poultry
Identify and remove bone, cartilage, plastics, wood, rubber, metal, or parasites. HSI can also measure the products chemical composition, including fat, protein and water content and meat tenderness.
Dried food – nuts and cereals
Detect discoloration or mold, identify and remove foreign objects like shell pieces, wood, stones and insects. Spectral imaging can identify different nut types such as almonds, cashew, walnuts, pistachio, peanuts.
Fruits berries and vegetables
Detect blemishes and bruising under the skin, measure ripeness and chemical quality, independent of the fruit color and size. It can also identify foreign materials such as wood, paper, metal or insects. Achieve better quality, ripe products with optimized shelf life, and reduce losses and waste.
Food packaging – heat seal inspection
Detect any contamination between the seal and the transparent or color printed package to identify if it is no longer airtight. As a result, air and contaminants such as molds, fungus or bacteria can enter the package and deteriorate the product well before the expiration date.
Seed and grain
Predict protein content, remove contamination, characterize the consistency and produce quality.
Recycling and waste management
The efficient recycling of waste into reusable raw materials is one of the significant efforts we must take to stop global warming and over-exploitation of natural resources.
A recycling plant must have sensors capable of separating a broad range of materials reliably and with high purity, and this is where spectral imaging can make a difference. The purer the recycled material, the higher its value. Efficient sensor-based sorting and recycling of different materials can be turned into profit with proper material handling methods and hyperspectral imaging technology.
Each plastic type has its specific spectral fingerprint which Specim FX17 cameras can identify with framerate and resolution that meets the requirements of industrial sorting processes. The spectral resolution is sufficient to identify and sort all different industrial plastic types, including PET, HDPE, PVC, LDPE, PP, PS, PE, and ABS, despite their color.
In 2019 SPECIM launched FX50 and quickly became the forerunner to solve the blacks plastics recycling problem as black polymers are in wide use in food, electronics and car industry.
FX50 is capable of detecting PS, PE, PP, ABS and PVC black plastics, rubber and specific additives in the MWIR region.
HSI can sort and separate plastics, textiles, metals, glass, paper, and cardboard.
Hazardous materials such as asbestos are always a concern. With hyperspectral imaging, it is possible to sort valuable and reusable materials such as concrete, wood, tiles, paper and cardboard, as well as different plastics safely, without exposing people to hazardous materials.
HSI can distinguish plant, animal, and artificial fibers. It shows good potential to help in solving the global challenge for managing textile waste and turning it to renewable materials.
Quality assurance of pharmaceutical products chemical composition with 100% inspection accuracy can be achieved with spectral cameras. Inspection is based on identification of products active ingredients. Products with correct looks but with wrong active ingredient or concentration can be identified in realtime
Thickness is a crucial quality parameter in thin films and coatings. The thickness and its homogeneity strongly affect the functionality of the film or coating, and they need to be monitored accurately.
Hyperspectral imaging significantly enhances the efficiency of current thin film and coating quality control systems based on optical spectroscopy. A line-scan hyperspectral camera allows 100% in-line inspection of thin films or coatings for increased consistent quality and reduced waste. In each line capture, spectroscopic data is produced with high spatial resolution over the full film width.
A spectral camera is a most accurate imaging technique to inspect and measure colors as it records the full visible spectrum in each image pixel.
Most current display panels and light sources are based on LED backlights. They produce inconsistent spectra and therefore their accurate color measurement is only possible by measuring the actual spectrum. Traditionally it is done by point spectrophotometers, which in production, limit the inspection to a few discrete points on the display surface, because of limited inspection time. Specim FX10 can provide a solution for 100% display surface testing at a production test station. Both luminance uniformity and color gamut can be inspected from the data.
Color control in digital printing
Digital printing is growing rapidly and is applied on paper, packaging, plastics, textiles, ceramics and metal (like cans) products. Printer control requires frequent accurate measurement of color test sheets and increasingly also real time in-line color measurement. Standard color cameras and multiband cameras are not accurate enough, and point-spectrometer based xy scanners are slow. Hyperspectral color scanner, based on SPECIM FX10 camera and intense LED illumination, will solve these issues, and the line scan concept can be transferred to in-line applications, too.
Mining & Oil
Spectral imaging offers mining and oil industries advanced solutions for more efficient exploration and higher automated mining and production processes with reduced waste streams and energy consumption.
Spectral imaging has been used for mineral exploration from the air over large rock surface areas for two decades. Now it is increasingly applied as a rapid method for reliable mineral analysis in drill cores and other geological samples. Spectral imaging in the infrared region reveals alteration patterns that can’t be seen in traditional logging by the human eye. Nearly all alteration minerals can be objectively and consistently identified and mapped in the infrared region. Specim SisuROCK scanner system has become a market-leading tool for analytical service providers, national geological surveys, and mining companies to rapidly image entire core trays and process large volumes of the core.
Mining & processing
Being able to map clays, talc, and other deleterious rock phases, spectral imaging can produce valuable information for building predictive models of mining and geometallurgical parameters. For this purpose, and in addition to core logging, spectral camera systems can be taken to the mining sites and installed on processing lines. Onboard a ground vehicle or a drone, the spectral camera can produce data for rapid mineralogical mapping of the entire mine wall faces.