Image Inspection Apparatus Using Closing Processing for Substrate Damage Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image inspection methods for substrates, such as liquid crystal substrates, face challenges in accurately detecting damage due to variations in substrate orientation, movement speed, and movement direction during conveyance by handling robots, leading to increased workload and inefficiencies in preparing reference images for pattern matching.
Innovation Solution
An image inspection apparatus and method that generate a reference image from a captured image of the substrate using image processing techniques like closing and pseudo edge detection, allowing for real-time inspection without pre-acquired reference images, and determining substrate quality by analyzing differences between the captured and reference images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional pattern matching inspection is used with pre-acquired reference images, then inspection accuracy can be maintained under stable conditions, but the workload increases considerably and the system becomes inefficient when conveyance conditions vary
Solution Approach 1:
The system uses the captured image itself to generate the reference image through closing processing, eliminating the need for external pre-acquired reference images. The image processing unit automatically creates a solid image reference from the captured substrate image, allowing the system to adapt to varying conveyance conditions without manual intervention or pre-prepared references.
Solution Approach 2:
The invention changes the parameter of reference image acquisition from static pre-acquired images to dynamic generation from captured images. By applying closing processing to the captured image, the system adapts the reference image to current conveyance conditions, maintaining inspection accuracy while improving efficiency.
2Adaptability or versatility
If multiple reference images are acquired under different conveyance conditions, then inspection accuracy under varying conditions can be maintained, but memory capacity requirements increase and processing complexity increases
Solution Approach 1:
The single reference image generation process serves multiple functions: it adapts to different conveyance conditions, provides a basis for pattern matching, and eliminates the need for multiple condition-specific reference images. The closing processing universally handles various image variations caused by different conveyance parameters.
Solution Approach 2:
The invention extracts only the essential substrate image information from the captured image through closing processing, removing unnecessary variations caused by conveyance conditions. This extracted solid image serves as the reference, simplifying the system while maintaining adaptability.
3Measurement precision
If image alignment processing is performed between captured images and reference images with varying shapes and sizes, then inspection can be performed, but the alignment processing becomes troublesome and time-consuming
Solution Approach 1:
The closing processing is performed in advance on the captured image to generate a solid reference image before pattern matching. This preliminary processing standardizes the reference image shape, eliminating the need for complex alignment operations during the inspection process and reducing processing time.
4Productivity
If pre-acquired reference images are stored for later use, then inspection speed can be improved, but the workload for acquiring and managing reference images increases considerably
Solution Approach 1:
The system generates its own reference image from the captured image through closing processing, eliminating the manual workload of acquiring and managing pre-acquired reference images. This self-service approach maintains fast inspection speed while removing the burden of reference image preparation.
Data Source
AI summary
An image inspection apparatus is provided with an image sensing unit, a determination unit, and a reference image generation unit. The image sensing unit captures an inspection target. The determination unit takes a difference between a reference image that includes a solid image of the inspection target and a captured image that is captured of the inspection target. The determination unit thereby extracts an image that is not included in an image of the inspection target in the reference image, and determines the quality of the inspection target based on the extracted image. The reference image generation unit generates the reference image that is used by the determination unit by performing predetermined image processing to change to an image showing the inspection target that is included in the captured image into a solid image.