Machine Vision
Pixel Accuracy Calculator
Convert between physical measurements and pixels using camera resolution, field of view, and working distance.
Machine vision workflowStep 3 of 6
- Size lens
- FOV specs
- 3Resolution
- Cell layout
- Focus
- Bandwidth
Next workflow step
Continue to Camera Placement Planner — 25.1 mm lens · 140×117 mm zone — drag layout in 3D
How it works: Scene width is 2 × WD × tan(HFOV/2). Pixel size is scene width divided by horizontal resolution. Object size in pixels divides physical size by pixel size.
Learn about this tool
What it is, when to use it, formulas, and common questions
What is this?
Converts between physical measurements (mm) and pixels using camera resolution, field of view, and working distance.
When do engineers use it?
Use when validating inspection resolution — e.g. confirming a 0.5 mm defect spans enough pixels for your algorithm.
Formula
mm/px ≈ scene width (mm) ÷ image width (px)
Example
At 0.05 mm/px, a 10-pixel feature represents roughly 0.5 mm — often the minimum for reliable edge detection.
FAQ
- How many pixels do I need across a defect?
- A common rule of thumb is 3–5 pixels for presence/absence and 10+ pixels for measurement or classification. Confirm with your algorithm vendor — lighting and contrast often matter more than the last pixel.
- Is mm/px the same as optical resolution?
- No. mm/px is geometric sampling. True resolving power also depends on lens MTF, aperture, lighting, and motion blur. Use this tool to check sampling, then verify optically.
- How is this different from the FOV calculator?
- FOV Calculator sizes the scene and lens. Pixel Accuracy takes a known scene width (or FOV) and resolution and tells you whether a physical feature spans enough pixels.
- Does working distance change mm/px?
- Yes, if FOV changes with distance. At a fixed focal length, moving farther away widens the scene and coarsens mm/px. Recalculate whenever mount height changes.