In the food industry, combinations of different machines and technologies can be used to prevent contamination by foreign materials. In order to sort out undesirable material, some machines use high pressure air jets while others use screens and vibrations. Each technique has different capabilities for sorting differences in material size, weight, and specific gravity.

More advanced techniques can sort differences by optical characteristics using cameras. These different
techniques may be used at different steps of the processing chain in order to prevent contamination in
the final product. Food manufacturers in Japan have had a particularly strong responsibility to produce high quality products after the Product Liability Law was enacted in 1995. As a result, there has been heavy investment in machinery which prevent food contamination and the variety of such machines available on the market has proliferated.


Optical sorters have evolved significantly
With the advancement of technology, optical sorters, which can sort material differences by optical characteristics, have evolved significantly. Not only are these machines able to use visible light to distinguish differences in appearance, they can also use near-infrared light to see differences in composition to identify foreign material which has a similar colour to the product.


Near-infrared light emitted by a material is well-known to provide a signature of the material's components. Spectroscopy is the technique used for optically analysing such light to determine these components. Generally, near-infrared light between 900-1700 nm is used for analysis. In an optical sorter utilising near-infrared light, digital cameras and image processing techniques are used to filter for the wavelength bands which are characteristic of different material components.

The strength of absorption at these wavelength bands, expressed as brightness and darkness, is thereby used to separate foreign materials from desired product. This identification method can sometimes be adversely affected by the colour characteristics of the material. If the desired product has strong absorption in wavelength bands in the near-infrared, the method is most effective when the colour of the foreign material is light coloured, like white or beige.

When the foreign material is darker in colour, like black or brown, the difference between the foreign material and the desired product can be difficult to identify.

High-quality sorts of varied materials
Foreign objects to be removed include stones, glass, plastic resin, wood, small animal bones, insects, and others. Because of the many different types of objects needing to be removed, it is often necessary to use multiple wavelength bands rather than just one. The downside of using many wavelengths is complexity.


Sensitivity adjustments become more difficult and the operator must be more skilled to make effective use of the system. This situation has led to the need for personnel skilled with adjustment to perform proper quality control measures after the machine is installed. Unfortunately, it is difficult to train and secure these individuals.

The latest model in Satake's NIRAMI series addresses these concerns by making it easier to do high-quality sorts of varied materials. High-end models include a dual-IR system which uses two different IR wavelength bands to handle a larger variety of foreign materials. NIRAMI is also equipped with Satake Smart Sensitivity and includes a newly added MIX mode capable of automatically creating sensitivity settings from five wavelength bands using its on-board AI technology.

Maximising yield & preventing contamination
Satake Smart Sensitivity is a system which maps optical signals to a 3D space. This space represents 16.78 million combinations of the principal colours red, green, and blue. The space is then algorithmically rotated in order to find the best discriminatory angle for identifying foreign material from good products. The system greatly simplifies the creation of sensitivity in a 3D space, a very difficult task for an operator. With the newly added MIX mode, it is now possible to select the most effective settings using red, green, blue, as well as the two new IR signals.

This space is 66,000 times larger than visible colour alone, representing over a trillion colour combinations. This allows for a much higher discrimination accuracy than that achievable by using visible or infrared wavelength bands alone. The sensitivity created by the MIX mode can also be adjusted with one parameter to change the strength of the sorting degree. This allows for both a high degree of discrimination and a high operator usability.

Up to six sensitivities can be registered at once and used simultaneously. By freely registering combinations of wavelength bands matching characteristics of different foreign materials, it is possible to create an optimal sensitivity for a given sorting task, maximising yield and preventing contamination.

Consumers demand higher safety & quality
Technological advances have made it possible to combine colour and component information to perform advanced sorting with high accuracy. Optical sorters can now remove a large variety of foreign materials as well as identify undesired product characteristics like damaged parts and mould. At the same time, advanced techniques and AI technology have made it possible to easily maintain and manage this high sorting accuracy. These factors are expected to become increasingly important as consumers demand higher safety and quality.

Note from the editor:
If you have enjoyed what you have read here, then why not treat yourself to a International Aquafeed magazine in-print subscription?

For a small fee, you too could join the many thousands of others who already receive content like this and more through their door very month.

To find out more about our very competitively priced monthly subscrition packages, visit our website by clicking on this LINK.

Article contributed by Nick Ikeda, Head of Digital Technology Research Office, Satake, Japan.

You might also like

Latest Videos

Leave A Comment

Don’t worry ! Your email address will not be published. Required fields are marked (*).





QR Code