Cracks on the bottom or edges of glass vials and ampoules and other defects such as scratches, bubbles, black spots, and inclusions can cause contamination of the contents or a breach of sterility. Vial and ampoule inspection must identify these defects before they reach customers to protect patient safety. Pharmaceutical glass can also exhibit minor defects, such as scratches, that do not affect function and can safely be shipped.
Given the transparent nature of the borosilicate glass used in vials and ampoules, it can be difficult to discern between defective and tolerable defects with standard machine vision, both because of transparency and reflections, and because rule-based vial defect detection cannot be programmed to handle all possible unpredictable defects.
SolutionBank Deep Learning is designed to solve these types of complex inspections since it is able to distinguish between true defects and acceptable cosmetic anomalies. The defect detection tool trains on a wide selection of glass vials, at different angles, to learn the full variation of normal parts, including the acceptable level of cosmetic defects. Then, as it scans through the vials it analyzes and flags any that are outside the acceptable range, while minimizing false positives caused by cosmetic defects.
For more details, business please contact:
SOLUTIONBANK CORPORATION
Headquarter: 9 Phan Ke Binh Street, Da Kao Ward, District 1, Ho Chi Minh City, Vietnam
Hotline: +8496 113 0541
Email: [email protected]
