Glass Bottle Inspection Using Differential Imaging and Adaptive Templates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Glass bottle manufacturing factories face challenges in detecting defects like checks in bottle-mouth portions due to variations in mold release agents and temperatures, leading to false judgments and the need for frequent updating of reference templates to maintain high-throughput inspection capabilities.
Innovation Solution
A method and apparatus that produce differential images from successive camera captures of glass bottles while rotating, comparing these images with a template to judge defect-free status, and combining them to create a provisional template for updating the reference image, ensuring accurate and rapid inspection and template correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a template matching process is used for inspecting glass bottles, then defect detection capability is improved, but the system requires frequent template updates due to mold variations, increasing operational complexity
Solution Approach 1:
The system performs preliminary actions by capturing and storing multiple reference images of defect-free bottles under various mold conditions before inspection begins. These pre-captured images serve as a comprehensive reference library that automatically adapts to mold variations, eliminating the need for frequent manual template updates while maintaining high defect detection accuracy
Solution Approach 2:
The system changes the parameter of reference image diversity by storing multiple images with different characteristics (captured at different times, under slightly different conditions) rather than using a single static template. This allows the inspection system to accommodate mold variations without requiring manual template reconfiguration
2Measurement precision
If multiple images are captured and processed to create accurate templates, then inspection accuracy is improved, but inspection speed decreases
Solution Approach 1:
Multiple reference images are captured and processed in advance to create a comprehensive reference library before production begins. This preliminary processing consolidates the computational workload, allowing rapid comparison during actual inspection without sacrificing accuracy
Solution Approach 2:
The system merges multiple reference images into a consolidated reference library that captures the full range of acceptable variations. This combined reference set enables faster single-pass comparison during inspection while maintaining the accuracy benefits of multiple reference images
3Measurement precision
If the inspection system adapts to mold variations by updating templates, then measurement accuracy is improved, but the inspection process time increases
Solution Approach 1:
The inspection system performs self-service by automatically capturing new reference images and updating the reference library without manual intervention. When mold variations are detected, the system autonomously adapts by incorporating new images into the reference library, maintaining accuracy while eliminating time-consuming manual template updates
Solution Approach 2:
The system implements feedback by continuously monitoring inspection results and automatically adjusting the reference library based on detected variations. This closed-loop approach allows the system to adapt to mold changes in real-time without manual intervention, maintaining high accuracy while minimizing downtime
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables high-speed inspection and template updating, allowing for efficient detection of defects and maintaining accurate reference images despite variations in mold conditions, supporting high-throughput glass bottle inspection processes.
Implementation Method 1
Scattered light emitted from the illuminating unit is applied to the bottle-mouth portion of the glass bottle, and if there is a check, the light is reflected by a crack plane of the check and is thus illuminated brightly.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a glass bottle inspection method and apparatus, and more particularly to a glass bottle inspection method and apparatus for detecting a defect at a specific location of a bottle-mouth portion and the like of a glass bottle by an imaging process. The glass bottle inspection method and apparatus performs successively producing differential images from original images successively captured from the glass bottle to be inspected while the glass bottle is being rotated about its own axis, comparing the differential images with the template to judge whether the glass bottle is defect-free or not, and combining all the differential images obtained from the glass bottle to be inspected in one inspection cycle to produce a differential composite image, using the differential composite image as a provisional template when all the differential images obtained from the glass bottle in one inspection cycle are judged as representing a defect-free glass bottle, and correcting the template using the provisional template.