Technology
Electro-Optical & Infrared Perception
Add visual context where visual evidence is available.
What it observes
It observes what a camera can show: shape, movement and, where the light and the conditions allow, a class of object. It adds visual context to a track that other sources have found.
The principle
A public-safe diagram of the mechanism, drawn without real values.
The diagram shows sources marked present or absent over a shared scene: an optical source contributes only where it has a view, and says nothing elsewhere. No frame or mark is drawn around any object.
Where it sits in the chain
It sits beside the other sources in the first band, and its output is joined in correlation and in classification and context. It refines a track; it does not start one on its own.
What it does to the data
An image becomes an observation with a class and a stated confidence, attached to the track it supports. A class that the image cannot support stays unknown.
How it shows up in the data
| Time (relative) | Wind | Visibility | Temperature |
|---|---|---|---|
| 1 | 12 | 60 | 8 |
| 2 | 14 | 58 | 9 |
| 3 | 18 | 52 | 11 |
| 4 | 22 | 40 | 14 |
| 5 | 20 | 36 | 17 |
| 6 | 16 | 42 | 18 |
| 7 | 14 | 50 | 17 |
| 8 | 13 | 56 | 15 |
| 9 | 15 | 60 | 12 |
| 10 | 17 | 62 | 10 |
Strengths / Constraints
Strengths
- It can tell an object's class where radio and radar sources cannot.
- Its output is easy for a person to read and to question.
- It adds context, such as the surroundings of a track, that other sources do not carry.
Constraints
- It sees only what is in view, in light and weather that allow it.
- A class drawn from an image is a candidate, not an identity.
- It depends on where the camera stands and where it points.
Environmental dependencies
Light, haze, rain, glare and the background change what an image can support. The page states the conditions that apply and gives no recognition rate.
Limitations
- Source availability: a camera is a single view; where it does not look it says nothing.
- Line of sight and environment: darkness, weather and obstruction remove the image, and silence then means no view, not no object.
- Confidence variability: a class that is clear in good light can be uncertain a minute later.
- Validation scope: the technology is a design at the evidence level of its projects; no recognition rate, range or error rate is stated.
Evidence and validation
This page claims nothing above the level of the projects it rests on.
- Electro-Optical AI Detection & ClassificationEvidence level: E0 · Concept
- Hyperlocal Atmospheric IntelligenceEvidence level: E0 · Concept