Button Battery Cosmetic Inspection
Detect a wide range of weld and other cosmetic defects on batteries
A breakthrough in complex inspection, part location, classification, and OCR
Button (or coin) batteries and cylindrical batteries are sealed by welding a cap onto a housing containing electrodes and electrolyte. A low-heat method such as laser or resistive welding is typical used to avoid damage to the battery contents but requires precision to ensure that the cap seal is secured. There is little margin for error. Defects can result in electrolyte leakage and decreased performance.
Welding defects can appear anywhere around the perimeter and can vary widely in appearance. Dents, scratches, and other defects can appear anywhere on the body of the battery. Both welding and cosmetic defects can be difficult to discern against the reflective surface of the cell battery. All of this makes it difficult for traditional machine vision to accurately detect the full range of defects found of a button battery.
And because these batteries are produced in high volume, and the defects are small and hard to spot, they present a similar challenge for manual inspection.
To identify battery weld defects, Cognex Deep Learning’s defect detection tool is trained on a wide selection of properly sealed batteries to learn the full variation of normal parts, including the acceptable level of cosmetic defects. As it scans through battery caps it analyzes and flags any welds that are outside of the acceptable range, while minimizing false positives from purely cosmetic defects.