Detecting scratches on shiny metal surfaces and transparent glass simultaneously is one of the most challenging machine vision problems. Reflections confuse traditional algorithms, scratches are often sub-pixel in width, and distinguishing real defects from surface artifacts requires advanced image analysis.
Multi-modal approach
We combine multi-angle lighting, polarised imaging, and Retina AI deep learning to reliably detect scratches on challenging surfaces. Our system adapts to different material types — polished stainless steel, anodized aluminium, painted surfaces, chrome plating, and transparent optical glass.
Applications
- Luxury watch components — case polishing verification, sapphire crystal inspection
- Optical devices — lens quality, mirror surfaces
- Consumer electronics — metal enclosures, screen protectors
- Automotive trim — chrome finishing, painted body panels
- Kitchenware — stainless steel utensils, glass cookware
Why 3HLE
Our Swiss engineering team has solved this problem for Haute Horlogerie clients where surface tolerances are measured in micrometres. The same technology applies to any industry requiring perfect finish quality.
Send us samples for a feasibility evaluation.