AOI Text Verification via Sub-window Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated Optical Inspection (AOI) machines face high false alarm rates due to strict inspection standards, leading to decreased production line efficiency, as they struggle to accurately identify IC components with varying font types, resulting in manual reconfirmation and standard adjustments.
Innovation Solution
An electronic device and image processing method that preprocess images to extract text areas, divide them into sub-windows, and compare these sub-windows with standard images to determine text similarity, marking images as qualifying or non-qualifying based on predetermined similarity values, thereby reducing errors and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If strict inspection standards are used in AOI machines, then measurement precision is improved, but false alarm rate increases and productivity decreases
Solution Approach 1:
The text area extraction process is segmented into multiple stages: initial text area identification, text window extraction with sub-windows, and hierarchical comparison levels. This segmentation allows the system to focus computational resources on critical text regions while ignoring irrelevant areas, maintaining high inspection accuracy without proportionally increasing processing time for all images.
Solution Approach 2:
The patent applies partial action by performing full text area extraction and comparison only on images that fail initial quick checks. The system first performs rapid whole-image comparison, and only for images that are uncertain or flagged does it proceed to the more time-consuming text area extraction and detailed sub-window comparison, thus maintaining productivity while ensuring measurement precision when needed.
2Measurement precision
If strict inspection standards are used to verify text on IC components, then measurement precision is improved, but device complexity increases due to font variation handling
Solution Approach 1:
The patent transforms the inspection approach by changing parameters from font-specific matching to structural feature matching. Instead of requiring exact font type, size, and style matches, the system extracts text areas and compares structural patterns, character arrangements, and relative positions. This parameter change allows the system to handle font variations from different manufacturers while maintaining verification accuracy, without requiring complex font libraries or manufacturer-specific configuration for each font type.
Solution Approach 2:
The system creates standardized text area templates from qualified IC components that represent different manufacturers and font styles. These templates capture the essential structural characteristics without being tied to specific fonts. When inspecting new components, the system compares against these copied templates rather than requiring exact font matches, simplifying the inspection standard management while maintaining precision across various font types.
3Measurement precision
If manual reconfirmation is performed for AOI-detected defects, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary text area extraction and comparison on all images before they reach manual inspection. By pre-processing images to identify and compare text regions automatically, the system filters out many false positives that would otherwise require manual reconfirmation. This preliminary automated text verification reduces the volume of images needing manual review while maintaining high detection accuracy, thereby reducing time loss without sacrificing precision.
Solution Approach 2:
The patent introduces an intermediary text area extraction and comparison system between the initial AOI detection and final manual reconfirmation. This intermediary layer automatically verifies text regions using the multi-level comparison method, acting as a filter that resolves many uncertainties before manual inspection. This intermediary process maintains measurement precision by catching text-related false positives early, while reducing time loss by minimizing the number of images that require time-consuming manual reconfirmation.
Data Source
AI summary
An image processing method for identifying text on production line components obtains an image to be recognized and a standard image for reference and extracts a first text area of the image to be recognized. A second text area of the standard image is obtained, and a text window is extracted based on the second text area. The method further obtains a target text area of the image to be recognized based on the first text area and the text window, and obtains a first set of first text sub-areas, and obtains a second set of second text sub-areas, by dividing the second text area into sub-windows of the text window. The method further marks the image to be recognized as a qualifying image when each first text sub-area of the first set is the same as a corresponding second text sub-area of the second set.