A glass bottle flaw recognition device and method based on double-ring type track sampling
The glass bottle defect identification device and method using dual-ring track sampling, by utilizing multi-view image acquisition and logically constrained neural networks, solves the problems of high false detection rate and liquid interference in glass bottle detection, and achieves efficient and automatic defect identification and sorting.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SICHUAN UNIVERSITY OF SCIENCE AND ENGINEERING
- Filing Date
- 2026-04-22
- Publication Date
- 2026-07-21
AI Technical Summary
Existing glass bottle defect detection technologies have a high false detection rate, significant liquid interference, cannot perform multi-view sampling, and cannot perform online high-speed detection of glass bottles containing liquid.
A glass bottle defect recognition device based on dual-ring track sampling is adopted. It utilizes multi-view image acquisition and a pre-trained glass bottle defect recognition model, combined with a logically constrained neural network and asynchronous servo drive, to eliminate liquid refraction interference and achieve multi-view collaborative sampling and high-speed detection.
It improves detection accuracy, reduces false detection rate, realizes automatic identification and sorting of glass bottle defects, reduces labor costs, and expands the practical application scope of detection.
Smart Images

Figure CN122090239B_ABST