Some of the richest sources of proteins – a staple of a healthy diet – are seafood, meat and poultry. But whether these are processed for human consumption or pet food, achieving food safety and product quality is not always easy. Which is why TOMRA offers application-specific sorting solutions for proteins.

Automated sorting machines for protein-rich foods (such as meat, nuts, grains, and dairy) typically utilize multispectral または ハイパースペクトル imaging technology. To achieve precise detection of protein content, moisture, and fat, and to effectively reject foreign materials, these sorters employ a combination of LED wavelengths ranging from visible light to 短波赤外線 (SWIR).

The following are the core wavelengths used in protein food sorting and their specific functions:

1. 短波赤外線 (SWIR, 1000nm – 2500nm) — The Core Chemical Fingerprinting Band

This is the most critical band for detecting protein, fat, and moisture, as organic molecules exhibit specific “fingerprint” absorption peaks within this range.

• 1180nm – 1220nm: The second overtone absorption band for fat and moisture. It is used to differentiate fat from lean tissue (the fat-to-lean ratio) in meat processing.
• 1450nm – 1500nm: A strong moisture absorption band. It is used to monitor freshness and dehydration levels, or to identify foreign objects based on differing water content.
• 1650nm – 1750nm (The Signature Wavelength): The primary absorption peak for protein (N-H bonds). This band directly senses the concentration of protein molecules and is considered the “signature wavelength” for protein sorting systems.
• 1940nm: The peak absorption point for water.
• 2050nm – 2180nm: The combination frequency band for proteins and amino acids. It is used for high-precision compositional analysis, such as protein-level grading in grains.

2. Near-Infrared (NIR, 700nm – 1000nm) — Basic Feature Recognition

• 850nm – 940nm: Standard infrared illumination. It is used to identify surface defects in seeds and nuts or to distinguish inorganic foreign materials, such as plastics and stones.

3. Visible Light (RGB, 400nm – 700nm) — Morphological and Aesthetic Sorting

• 450nm (Blue), 525nm (Green), 630nm (Red): Used to detect anomalies in shape, size, and color, such as meat discoloration or mold in nuts.

4. Ultraviolet (紫外線, 365nm – 395nm) — Fluorescence Detection

• 365nm: Used to excite fluorescence in aflatoxins (primarily in nuts and grains) or bone fragments. This allows the system to eliminate carcinogens and bone residue that are otherwise invisible to the naked eye.