The key point
Resolution only becomes useful when you combine it with field of view, target size and distance. At the same 50 m distance, a narrower lens can put more pixels on a small target while surveying less area per image.
The question behind “is 640 enough?”
An inspection team needs to know whether a small defect will occupy enough of the image. A wide-area observer needs to know how much ground fits in one view. Neither question can be answered by detector resolution alone. Write down your smallest relevant target, camera-to-target distance and the task: detect a difference, recognise a feature or measure a temperature.
Native 640 × 512 has twice as many pixels across as native 320 × 256 and four times the pixel positions overall. At the same field of view and distance, a target spans twice as many pixels across. That arithmetic does not promise four times the usable range or temperature accuracy.
Calculate the pixels on your target
For a flat target plane perpendicular to the view, scene width = 2 × distance × tan(horizontal field of view ÷ 2). Approximate width per pixel = scene width ÷ native horizontal pixels. Target width in pixels = target width ÷ width per pixel. Use degrees for the field of view when entering the tangent in a calculator.
These are geometric estimates. Use the camera-to-target distance, not simply flight altitude when viewing obliquely. Sloped surfaces, off-axis targets, lens distortion, motion and processing require further evaluation.
The 32.9° and 22.8° views are the catalog’s listed horizontal fields of view, with a stated tolerance of ±5%. Distances and target size in the table are examples. The 320-pixel case is a hypothetical comparison, not a fourth catalog product. Values are rounded and are not measured detection performance.
| Native width / horizontal view / distance | Approximate scene width | Approximate target pixels across |
|---|---|---|
| 640 pixels / 32.9° / 50 m | 29.5 m | 4.3 pixels |
| 640 pixels / 22.8° / 50 m | 20.2 m | 6.3 pixels |
| 320 pixels / 32.9° / 50 m | 29.5 m | 2.2 pixels |
| 640 pixels / 32.9° / 100 m | 59.1 m | 2.2 pixels |
Choose coverage or detail deliberately
At 50 m, the narrower view in the example covers about 20.2 m instead of 29.5 m across. A 0.20 m target spans more pixels, but a site survey may need more images to cover the same width. Actual route time also depends on the vertical view, overlap, orientation, flight speed and capture rate.
If the target is too small in the required safe working geometry, evaluate a different lens, native resolution or capture position. Digital zoom enlarges the samples already recorded. It does not turn a target spanning about 2 pixels into a newly resolved feature.
Detection, identification and measurement are separate tests
A small warm object may be visible even when its shape is not recognisable. Measuring its temperature is another requirement: the measurement area can include background and the camera’s measurement spot may require multiple detector pixels. Consult the model’s documented measurement field of view or spot-size guidance instead of treating a single pixel as an accurate thermometer.
FLIR’s distance-to-size explanation illustrates why target size and distance belong together in measurement planning. There is no universal pixel count in this article that certifies detection, identification or measurement for every target. Verify the required outcome with representative original files.
Do not compare an enlarged stream with a native detector
A cited Teledyne FLIR module announcement describes native 1280 × 1024 and an 8 µm pixel pitch. The catalog payloads here use native 640 × 512. An optional 1280 × 1024 processed output from a catalog payload is a different specification from that native detector announcement.
Pixel pitch also needs optical context. A smaller pitch does not automatically produce a narrower field of view: focal length and total detector size matter. For practical buying decisions, request the actual quoted lens and field of view rather than guessing from pitch alone.
Ask for a comparison that answers your job
Use the same target dimensions, distance, viewing angle and relevant scene conditions for each candidate. Request original files and identify whether they are native, cropped, enlarged or digitally zoomed. Include the real radio link and display if the job relies on live imagery.
- Record target width, distance, horizontal field of view and estimated pixels across before the demonstration.
- Compare the actual decision: locate the target, identify the feature or obtain a usable measurement.
- Keep original images, lens details, settings and observed failures with the evaluation.
- Choose the configuration that meets the task with workable coverage, mass and integration cost.
Original source
View the original post and its available photos or video. Media stays hosted by X.
The original post could not load. Use the source link above.
Loading the embedded post connects your browser to X. You can read the article without loading it.
Sources & further reading
Match the payload to your requirements.
Compare documented specifications, then confirm the details of your integration.