When to use 2D vision or 3D measurement
Start with the acceptance rule. If a defect changes the image or outline consistently, evaluate 2D vision. If you reject by height, depth or profile, evaluate a method that measures that dimension.
Start with the defect and its acceptance rule
A scratch can be a visual defect, a geometric defect, or both. Ask quality which observable determines rejection. For a machined bore, compare “visible burr on the edge” with “burr exceeds a height limit.” The former can justify a lighting and 2D imaging trial. The latter calls for a measurable height reference.
Check what the sensor can actually see
2D performance depends on lighting, field of view, optics and the smallest relevant feature. For laser profiling, check line of sight, reference distance, measurement range and part motion. A narrow bore may hide the edge from the sensor. Change the viewing geometry or test another measurement method before committing to equipment.
Plan a test that would change the decision
Run representative known-good and known-bad samples under expected production conditions. Record missed defects and false rejects. Include variation in finish, position and speed. For 3D, check repeatability against the acceptance tolerance. If the result is inconclusive, record the cause before adding another sensor.
Read published specifications in context
The 12 MP rating of an industrial smart camera describes the sensor, while the 55 mm reference distance of a line profiler describes its mounting geometry. Neither determines the smallest detectable burr on your workpiece. The example review shows how MentiVision turns those specifications into a test plan.