Guide · Running the line
How do you review unverified eggs by hand?
An unverified egg is one that Vigiovo cannot measure in full: it touches the image border, overlaps another egg or has an abnormal shape. No such egg receives a grade. Manual review checks each one by hand, records the reason and uses the count of each cause to correct spacing and lighting.
Updated · 5 min read
Why do some eggs stay unverified?
An egg stays unverified when Vigiovo cannot measure it reliably. The rule is fixed and applies to each egg in the image: the egg is marked unverified, and never graded, when it:
- touches the image border, so part of it is outside the frame;
- has an area outside 0.35 to 2.5 times that of a typical egg;
- has solidity below 0.90, a concave shape that signals overlap;
- does not look like an ellipse or has an axis ratio above 1.8.
| Cause | Rule that triggers | Practical correction |
|---|---|---|
| Egg at the border | touches the image border | Adjust the framing so the conveyor fits in the frame |
| Egg touching another | solidity below 0.90 or area outside the range | Increase the gap between eggs on the conveyor |
| Egg with an odd shape | does not look like an ellipse, or axis ratio above 1.8 | Check by hand and review the camera angle and framing |
In practice, the rules cover three causes: the egg at the border, the egg touching another egg and the egg with an odd shape. Each one calls for a different correction on the line, so the cause matters more than the total.
How do you find the cause of each unverified egg?
Each egg record carries a readable reason, together with the class and the measurements. To investigate, read the reason for each unverified egg in the shift, or count the reasons instead of only counting eggs. The reason that dominates the count points to where the line has a problem, and manual review shows whether the recorded reason is the real one.
A total alone is not enough. Two shifts with the same count can have different causes: one with eggs touching the border, another with overlapping eggs. The fixes are different, so split the count by reason before acting. The format of the records is described in per-egg records for audits.
How do you set up a review point on the line?
Create a fixed point where unverified eggs are separated and checked by a trained person, always with the same manual grading rule. Without a fixed point, unverified eggs get lost in the flow and the count stops representing the shift.
Set up the point in this order:
- Separate the unverified eggs into their own container, apart from the graded ones.
- Check each egg by hand, using the same class reading that the training teaches.
- Record the reason and the class you would give, to compare later with the system.
- Add up the shift's reviews by reason and pass the totals to the quality team.
The class reading used at the review point should match the operator training. The point is also useful as a measurement during the camera pilot.
How do you use the count to adjust spacing and lighting?
Fix the dominant cause before touching any limit. If many eggs touch the border, the problem is the framing, and no limit changes that. If many eggs show up touching each other, the gap between them on the conveyor is the thing to review. If shadow or glare blurs the outline of the egg, lighting is the point to review, as lighting for egg inspection explains.
Changing the limits to lower the count of unverified eggs trades a visible result for an invisible error. The area and solidity limits exist so that an overlapping egg is not measured as if it were a whole egg. After each change of spacing or light, count again over a comparable shift and compare the reasons.
When can manual review be reduced?
Manual review can be reduced when the dominant cause falls, not when the total falls by chance. Compare the reasons shift by shift after each change in spacing or light. If the cause disappears and the total drops with it, the fix worked; if the total drops but the cause remains, the change only hid the problem.
Before reducing the review, keep the point running for a few shifts and confirm that the cause stays low. The routine for records and review can be tested in your line pilot before it becomes a permanent rule.
Frequently asked questions
Is an unverified egg a bad egg?
- Not necessarily. It means Vigiovo cannot measure the whole egg, because it is at the border, touching another egg or has an odd shape, so it gives no grade. It needs a manual review before any decision is made.
Can I raise the limits to get fewer unverified eggs?
- Not without weighing the risk. The area and solidity limits protect against measuring overlapping or incomplete eggs. Loosening them lowers the count, but the error becomes invisible. Fix the spacing, the framing and the lighting first.
How do I find the reason for each unverified egg?
- Each egg record carries a readable reason together with the class and the measurements. Read that reason for each egg you review, or count the reasons across a shift to see which cause dominates. That way the fix goes to the right cause, not to the total.
Sources
How Vigiovo helps
Vigiovo grades eggs on the conveyor as clean, dirty, cracked or unverified, in real time, with deterministic computer vision: every decision comes with the measurements behind it, on an ordinary CPU.