Surface Inspection by Deep Learning
Case • 16.08.2022
Finding and implementing optimum automated product surface inspection technology, which is more accurate, repeatable and consistent than human quality inspection, while maintaining flexible to product variation and future adaptability.
Technology highlights
Medical Product
• Automated optical inspection of various surfaces
• 100% in-line inspection
• Optical system for 3D inspection, including rotation of product
• Deep learning algorithms used for qualification
• Searching for defects: damage, burrs, dents, scratches, color differences, burning effects
• User can easily access and program system for a wide range of recipes for combining camera inputs, deep learning and quality criteria
Application specific benefits
• Algorithms use Deep Learning Technology
• Training of known and unknown product defects (supervised/unsupervised)
• Wide range of product portfolio (>1000 product types possible)
• Replaces inspection by humans: more reproducible
