Quality you can see, and quality you cannot. We evaluate food quality comprehensively with 2D, 3D and AI visual inspection and near-infrared internal quality measurement, and cover everything from inspection to NG discharge and integration into your existing line.
Illustrative animation. The values and grades shown are for illustration only and are not product specifications or actual measurements.
Proposals that combine imaging, 3D and near-infrared
From inspection to NG discharge and integration into existing lines
Applications in agriculture, processed food and fisheries
Optimized from top-and-bottom, all-face and sheet inspection
External quality inspection
Image inspection combining 2D, 3D and AI evaluates the quality you can see.
See 2D, 3D & AI
Internal quality measurement
Measures internal quality without cutting, turning the quality you cannot see into numbers.
See the Imes 10Appearance plus interior:comprehensive quality evaluation
Quality you can see, and quality you cannot. NIRECO evaluates the value of food accurately from both sides.
Fast, high-precision quality evaluation that improves shipping efficiency and stabilizes quality on sorting lines
What we inspect
Measure color
Detect surface blemishes
Measure shape and size
Detect dents and cracks
Multi-angle inspection that stabilizes quality and keeps foreign matter and defective products from reaching the market
Food external quality inspection systemDorayaki quality inspection system
Sheet product quality inspection system
What we inspect
Detect uneven browning and scorching
Detect cracks and chips
Detect shape defects
Detect adhering foreign matter
Automated species sorting and fat content evaluation that save labor and add value
What we inspect
Identify and sort species
Measure fish size
Measure fat content
Inspects the top and bottom faces at the same time
Example:Two-face (top and bottom) inspection system
Inspects all six faces: top, bottom and four sides
Example:Six-face inspection system
From cut sheets to roll-to-roll
Example:Sheet-form food inspection system
* The system configurations and inspection layouts shown are reference examples. Actual configurations vary with your required specifications.
When the defect is the same color as the sound area, conventional image processing struggled to detect it.
Cracked tomatoes, peeling tofu
Merely detecting damage is of little use for sorting by disease or for advising growers. What is needed is a judgment that classifies the type of damage.
Diseases on melons and mandarins
With foods of irregular shape and size, overlapping and touching items make it hard to separate outlines. Reflective cans were also a source of miscounts with conventional processing.
Counting touching food items or cans
High-precision measurement backed by decades of image processing and sensing technology.
Quantitatively evaluates color, shape, area and more, as well as adhering foreign matter.
Measures height, unevenness, volume, cross-section shape and more that 2D struggles to capture.
A laser line captures height data line by line to generate 3D image data of the object, measuring even slight surface unevenness and defects with high precision. Adding 3D inspection to the color and blemish detection of 2D image inspection captures fine unevenness that is hard to spot by eye.
3D is little affected by lighting conditions, so it measures the volume and shape of soil-covered potatoes even on a black conveyor. An integrated housing and a single control PC also keep the footprint small.
Identifies quality differences that conventional methods could not, with high accuracy.
AI judges appearance defects that image processing alone struggled to identify: same-color defects, complex shapes and defects with large variation from piece to piece.
Our long experience lets us condition images so the AI can pick out features easily before inference. We tune the model for each fruit and use XAI heat maps to check where it is looking, feeding that back into retraining.
Measures internal quality such as sugar content and acidity without cutting.
A near-infrared internal quality sensor converts differences in the absorption of light transmitted through the object into estimated quality values using a calibration curve built in advance. Sugar content and acidity of fruit are measured by the Imes 10 internal quality sensor, and crude fat in fish by the QG100 fish quality inspection system, which applies the same near-infrared analysis. Combined with visual inspection, it evaluates the quality you can see and the quality you cannot on the same line.
Internal disorders are graded by level for sorting against your quality standards. From fruit Brix measurement to fish fat content, we build the calibration curve for your target.
Examples of inspection items
Actual images captured by the inspection systems' cameras, with the inference and judgment results.
Original image
AI inference
Original image
AI inference
Original image
AI inference
* Values shown are the results of individual trials and installations, not guaranteed values. Accuracy varies with the target, imaging conditions, training data and calibration curve.
MelonsAI image inspection
Classes the AI sorts into
* The disorders that can be classified and the judgment accuracy vary with training data and imaging conditions.
Chinese yam3D inspection
* Weight estimation accuracy varies with the shape of the target and measurement conditions.
OnionsNear-infrared internal quality inspection
* Levels are the reference used when building the calibration curve. Judgment accuracy and thresholds vary with the target and calibration curve.
Tell us the items and defects you want to inspect, and your current line conditions.
Image processing + AI / 3D measurement (optional)
Cameras capture images of food on the conveyor line, and high-speed image processing evaluates each item against inspection criteria based on your specifications.
Measures images, height and internal quality, and merges data from the external and internal sensors.
Compares the measurements with quality standards to determine grade and size class.
Sorts in sync with conveying. Settings, measurements and outlet history can all be reviewed.
Links to sorting slips, databases and JA (agricultural cooperative) systems, managing everything from receiving to shipping.
At agricultural cooperative facilities and merchants nationwide
300sites and more
Installations of the AP8000 series external quality sensor
* Camera configurations are custom-designed for the target and specifications. Multi-lane discharge control is provided in units of 5 or 6 lanes.
Citrus, tomatoes, apples, melons, Chinese yam and more. Custom-designed for the natural variation in shape and color of fresh produce.
Fried tofu, thick fried tofu, udon, rice crackers, cookies, financiers and more. Custom-made to suit the product and line.
Crude fat of Pacific saury, mackerel and horse mackerel, species sorting, and color inspection of herring roe and salmon fillets, among others.
Examples of inspection targets
Specification values are representative values, not guaranteed values. Contact us about differences between inspection targets.
