# Human or automatic visual inspection: which should you use for egg grading? > Human inspection recognises unusual cases with flexibility, but it tires and varies over the shift. A camera applies the same criteria to every egg and records the reason for each decision, but it only recognises the defects its rules describe. In practice the camera does the continuous sorting, and people assess the unverified eggs and candling. Updated: 2026-10-09 · URL: https://vigiovo.com/en/resources/human-vs-camera-inspection · Português: https://vigiovo.com/recursos/inspecao-humana-vs-camera.md ## Key takeaways - A camera applies the same criteria to every egg in a shift; human inspection varies with fatigue and volume. - Each camera decision is recorded with its class, reason and measurements; human inspection usually leaves no per-egg record. - People judge unusual cases with flexibility; a camera only recognises the defects its rules describe. - A camera can't see inside the egg, so candling remains mandatory in registered Brazilian establishments (RIISPOA, art. 223). ## What does human inspection do well? Human inspection is flexible. An experienced operator notices a shell that looks wrong, decides on a case that doesn't fit the pattern and senses that something is off before being able to say why. That judgement is the person's main strength on the line. RIISPOA requires establishments to carry out a general assessment of shell cleanliness and integrity, and candling (art. 223). Those steps demand close looking, and this is where operator experience counts. The limit is well known: attention drops with fatigue, and two people can judge the same egg differently. That variation is not a failing of the person. It is what happens to anyone repeating the same task for hours. ## What does a camera do better? A camera applies the same criteria to every egg. For the same image the result is always the same, from the first egg of the shift to the last. It doesn't tire, and it doesn't change its standard when the volume rises. It also leaves a record. For each egg, [Vigiovo](https://vigiovo.com/en/how-it-works) produces a class, a readable reason and the measurements behind it. That makes a decision auditable later and helps trace where dirty and cracked eggs come from. Human inspection rarely leaves that trail. The camera's limit is what it can see. A camera over the conveyor can't see inside the egg, and Vigiovo only recognises the defects its rules describe. When an egg can't be assessed in full, because it touches the image border or overlaps another egg, the system marks it as [unverified](https://vigiovo.com/en/resources/glossary#unverified-egg) instead of guessing a class. ## Human or camera inspection: what does each one see and record? The table compares both methods on the same criteria. The camera column describes Vigiovo today. **Human inspection compared with a camera** | Criterion | Human inspection | Camera (Vigiovo) | | --- | --- | --- | | Consistency over the shift | Varies with fatigue and volume | Same criteria, same image, same result, always | | Record of each decision | Usually not | Yes: class, reason and measurements for each egg | | New or rare defects | Recognised and judged with flexibility | Only those its rules describe | | Ambiguous cases | Judged by experience, may differ between people | Marked as unverified | | Inside of the egg | Only with [candling](https://vigiovo.com/en/resources/glossary#candling) | Not visible | | Speed | Limited by the operator | About 90 frames per second on one CPU core | | Adjusting the criteria | Training and team review | Thresholds tuned at calibration with labelled eggs | ## How should work be split between people and the camera? The split that works best leaves repetitive work to the camera and judgement to people: 1. Let the camera do the continuous sorting of dirty and visibly cracked eggs, egg by egg. 2. Set aside the eggs marked unverified and assess them separately, by candling or human inspection. 3. Keep candling at the points RIISPOA requires, since it sees the interior that the camera cannot reach. 4. Review a sample of the eggs the camera passed to confirm the standard has not drifted. 5. Use the camera's records to compare batches and shifts, and investigate the cause when the dirty or cracked rate changes. This split uses what each method does best. For how candling and a camera combine on one line, see [candling or computer vision](https://vigiovo.com/en/resources/candling-vs-computer-vision). ## When does the camera need calibrating and reviewing? The camera must be calibrated with labelled eggs from your own line before it goes into use. Calibration should be repeated when the lighting, the conveyor or the camera changes. Your team labels those eggs, which also takes human judgement: the camera's reference is the decision the people on the line make. The step-by-step guide is [calibrating with labelled eggs](https://vigiovo.com/en/resources/calibrating-with-labelled-eggs). ## Frequently asked questions ### Can a camera replace human inspection? Not completely. The camera does continuous sorting with a fixed standard, but it can't see inside the egg and only recognises the defects its rules describe. Unverified eggs, and the candling that the law requires, still need a person or other equipment. ### Why does human inspection vary over a shift? Attention and precision drop with fatigue and with the number of eggs, and two people can judge the same egg differently. This happens with any repetitive task. A camera doesn't tire and keeps the same standard from the start of the shift to the end. ### Does a camera leave a record for each egg? Yes. For each egg Vigiovo produces a record with the class, the scores, a readable reason and the raw measurements. The record supports audits and helps find the origin of dirty and cracked eggs in production. ## Sources - [Decreto nº 9.013/2017 (RIISPOA)](https://www.planalto.gov.br/ccivil_03/_ato2015-2018/2017/decreto/d9013.htm), Presidência da República (Brasil) - [Como o Vigiovo classifica ovos sem IA / How Vigiovo grades eggs without AI](https://vigiovo.com/como-funciona), Vigiovo ## Read next - [Calibrating with labelled eggs, step by step](https://vigiovo.com/en/resources/calibrating-with-labelled-eggs.md) (Guide, 2026-10-09): A step-by-step guide to calibrating Vigiovo with a few hundred labelled eggs: what to label, which limits to tune and how close each decision sits to its threshold. - [Deterministic vision vs neural networks for egg grading](https://vigiovo.com/en/resources/deterministic-vision-vs-neural-networks.md) (Comparison, 2026-10-09): Deterministic vision and neural networks compared for egg grading: repeatability, explaining decisions, data needs, hardware and the limits of each method. - [Candling vs computer vision: what each one detects](https://vigiovo.com/en/resources/candling-vs-computer-vision.md) (Comparison, 2026-10-09): Candling compared with camera inspection in egg grading: which defects each detects, speed, record keeping and why the two methods work best together. ## 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. https://vigiovo.com/en/how-it-works.md 4 min read.