Scratch detection systems use a controlled view of a surface to identify marks outside an agreed acceptance limit. Define the unacceptable scratch and normal surface variation first. Then test whether lighting, optics and presentation expose that difference consistently across the required inspection area.
Speak directly with Oxford Vision Systems about automated surface-scratch detection.
Supply real surfaces with acceptable marks and scratches at the rejection boundary, including their direction and location. Trial the proposed lighting, viewing angle and presentation across normal finish variation. Record results against the agreed sample set and identify any areas or defect conditions outside the demonstrated coverage.
A scratch dimension alone is not a complete inspection specification. Its visibility changes with the surface and viewing arrangement.
Provide real examples at the acceptance boundary with their location, direction and appearance recorded. Separate cosmetic criteria from any functional requirement. A camera result should be compared with an agreed reference decision, not a changing personal judgement.
Include textures, colours, machining marks, coatings and production-batch variation that must pass. Keep genuinely marginal examples in the trial. Without these, a setup may detect obvious defects while rejecting normal products in routine production.
Compare lighting and viewing arrangements on the same sample set. Check scratches in different directions and positions, including edges and curved areas. Record blind or weak areas explicitly rather than assuming one successful image proves complete surface coverage.
Define rotation, tilt, position range and movement at image capture. Include fixtures, conveyor vibration and handling variation. The available view and repeatability of presentation must be assessed before choosing resolution or inspection timing.
Tie the inspection settings to the correct product and surface finish. Define part tracking, failed-part containment and the response to a missing image. Keep uncertain results separate until the agreed review or rework decision is complete.
Count missed defects and false rejects separately against the agreed sample decisions. Include repeated passes and normal presentation changes. Record the demonstrated defect range and surface coverage; do not turn a limited trial into an unrestricted detection claim.
Read how to define the scratch decision before specifying resolution to compare the practical trade-offs before agreeing an inspection or automation concept.
The right choice changes with visible defect or measurement definition for automated surface-scratch detection, good, bad and marginal sample range, your line layout and required results.
Create an indexed set of good, bad and borderline surfaces before adjusting the inspection. Retain the expected result, defect location and relevant finish for each item. Compare candidate optical arrangements on that same set, then confirm timing and handling on the line. Keep excluded areas and uncertain defect classes visible in the final scope.
Use the selection guide to compare the scope, then send the specification checklist with your application enquiry.
Your final solution for automated surface-scratch detection must reflect the law, current standards and site-specific risks that apply to your installation.