Work out vision system pixel resolution instantly with clear inputs, formula shown and shareable results.
Spatial resolution is field of view divided by sensor pixels. Detecting a feature reliably needs several pixels across it — two is the theoretical minimum, three to five is normal for measurement, and ten or more for accurate edge fitting or OCR.
Vision resolution
mm per pixel = FOV width / sensor pixels; pixels on feature = feature size / mm per pixel
Sampling theory needs at least two, and in practice edges fall between pixels, lighting varies and lens MTF blurs detail, so three or more gives repeatable detection.
Not necessarily. Reducing the field of view, using a better lens or telecentric optics often gives more usable resolution for less money and data.