How does optical sorting work?
Optical sorting works by imaging each item as it passes a camera and firing an air jet to remove the ones that fail. Product moves on a belt, on rollers or in free fall past colour, NIR or hyperspectral cameras. A model classifies each item in milliseconds, and a valve bank timed to the item's position ejects rejects into a separate stream. AI adds defect classes, such as bruising under the skin, that fixed colour thresholds handle poorly.
What does fruit sorting software do?
Fruit sorting software turns camera images into a grade for each fruit. It measures size, shape and colour, classifies blemishes, bruising and rot, and sends each item to an outlet lane by class. With AI, the grading rules are learned from labelled examples of your varieties, so a new variety or season is added by labelling images rather than rewriting thresholds.
What are examples of hyperspectral imaging in food?
Common examples of hyperspectral imaging in food are bruise detection in apples, fat and lean measurement in meat, nut and grain sorting, and foreign-material detection on produce lines. The camera records many narrow wavelength bands, so materials that look alike in colour can be told apart by their spectrum. On production lines, a few selected NIR bands often give most of the benefit at lower cost.
Is fish bone detection done with X-ray?
Yes, fish bone detection is done with X-ray, because pin bones sit inside the fillet where cameras cannot see them. Opsio's cameras handle the visible defects on the same line: blood spots, trim, fillet shape and fat lines. In a combined system, the X-ray result and the camera result are merged into one reject decision per fillet. The X-ray remains your bone control and keeps its own validation.
What cannot be detected by X-ray?
X-ray cannot reliably detect low-density foreign bodies such as plastic film, paper, cardboard, wood, hair and insects, because they absorb about as much as the food around them. It also does not see surface colour, blemishes or blood spots. Cameras can find many of these when they are on the visible surface, which is why food plants often run cameras and X-ray together, each for the hazards it suits.
What are double seam defects?
Double seam defects are faults in the folded joint between a can body and its lid. Visible ones include droops, vees, cut-overs, sharp seams and wrinkles; others, such as short body hook or low overlap, need teardown or seam scanning to measure. A camera after the seamer catches the visible faults on every can and points to the seamer station that produced them, while lab teardown remains your seam control.
What is a cap inspection?
A cap inspection checks that every bottle or jar has its closure fitted correctly. The camera looks for missing caps, cocked or high caps, broken or missing tamper bands and, on some lines, the right cap colour for the product. It runs after the capper, usually with two to four cameras or mirrors for a 360° view, and rejects faulty containers before labelling or packing.
How does fill level inspection work?
Fill level inspection measures how high the liquid stands in each container and rejects under- and over-fills. In clear glass or PET, a camera in front of a backlight finds the meniscus edge. In cans and opaque containers, an X-ray beam measures the level through the wall. The result is compared with the target for that product, and a pusher removes the container if it fails.
What is empty bottle inspection?
Empty bottle inspection checks containers for damage and debris before they are filled. Cameras view the base over a backlight, the sidewall from several angles and the finish (the threaded neck) from above, finding chips, cracks, glass fragments and residue. It matters most for returnable glass, where bottles come back with foreign objects, and it removes a hazard before product goes into the container.
What does label inspection software check?
Label inspection software checks that the right label is on the container and applied correctly. It verifies presence, position, skew, wrinkles and flags, reads the best-before date and lot code, and compares the label with the product running. Matching the label to the batch from your ERP is the practical check against allergen and label-claim mix-ups, a common cause of food recalls.
Is HACCP QA or QC?
HACCP is a preventive food safety system, so it sits under quality assurance, while the checks it relies on are quality control. A camera station is a QC check that produces records. It can support HACCP monitoring and verification, but it only becomes part of a critical control point if your HACCP team decides so, validates it and defines the critical limits, corrective actions and records for it.
What are the 7 basic HACCP principles?
The seven HACCP principles are: conduct a hazard analysis, determine critical control points, set critical limits, establish monitoring, define corrective actions, establish verification and keep records. Vision inspection data fits mainly into monitoring, verification and records: per-item decisions, reject counts, images and model versions, timestamped by batch. It supports these principles; it does not replace your metal detection or X-ray critical control points.
Is it safe for food to go through X-ray?
Yes, food that passes an X-ray inspection system is safe to eat. The dose used for inspection is very low, it does not make food radioactive, and regulators such as the US FDA set limits for X-ray inspection of food. Opsio's cameras do not use X-ray; where bone or dense foreign-body detection is needed, an X-ray unit runs alongside the cameras as a combined system.
