Overview Deep Learning automates regulatory compliance by doing multi-point inspections on blood testing kits. Our cameras pick up tough to spot bubbles in clear plastic, scratches in the casing, and subtle misalignment between layers.
Defects are rare and varied making a traditional vision very expensive and hard to get high accuracy. Current human quality control specialists are inconsistent and slow.
The Overview Solution
Our Deep Learning platform replaces human inspectors and can spot rare defects without ever having seen prior examples with near 100% accuracy.
How does the
Deep Learning Work?
Unsupervised learning quickly identifies large anomalies using no annotated data, such as scratches, large bubbles, and layer misalignments
Region-of-interest by region-of-interest inspection of each relevant spot finds all bubbles and divots
An overall generalized scratch detector flags all small to medium scratches
Reduced part rejection, increased yield, and effortless regulatory compliance.
Demo the Snap Platform to see how we improve your factory efficiency