Circuit Board Character Recognition Using Local Threshold Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The production of packaged modules in electronics is hindered by defects caused by impurities, scratches, and molding flashing, leading to monetary and resource losses, as well as reduced yield due to the need to identify and remove defective modules from production lines.
Innovation Solution
A circuit board processing system that uses a character recognition system to analyze digital images of characters on panels, applying local and global threshold values, along with contrast equalization and fragmented search regions, to classify and read characters effectively, even when damaged, thereby reducing processing time and increasing accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional global threshold value methods are used for character recognition, then the processing approach is simple, but the processing time is long and accuracy is reduced for damaged characters
Solution Approach 1:
The patent divides the character recognition process into multiple processing stages: initial global thresholding to identify candidate characters, followed by local thresholding on fragmented search regions to precisely extract damaged characters. This segmentation allows the system to handle complex damaged character cases without applying complex algorithms to all characters, thus reducing overall processing time while managing complexity.
Solution Approach 2:
The patent applies different thresholding strategies to different regions of the image based on local characteristics. Global threshold values are used for intact characters, while local adaptive threshold values are applied specifically to regions containing damaged characters. This local quality approach optimizes processing efficiency by applying complex algorithms only where necessary.
2Measurement precision
If traditional image analysis methods are used, then the system is simpler to implement, but the ability to read damaged characters is reduced
Solution Approach 1:
The patent performs preliminary global thresholding and character localization before applying more sophisticated local thresholding methods. By first identifying candidate character regions using simple global thresholding, the system prepares the data structure and search regions in advance, enabling subsequent accurate extraction of damaged characters without redundant processing.
Solution Approach 2:
The patent applies local thresholding and fragmented search region analysis only to specific areas where damaged characters are detected, rather than processing the entire image with complex algorithms. This partial action approach maintains high accuracy for damaged characters while avoiding the excessive computational complexity that would result from applying the same sophisticated methods to all characters.
3Productivity
If production lines continuously operate without interruption, then productivity is high, but defective modules with molding flashing cannot be identified and removed
Solution Approach 1:
The patent introduces an automated optical inspection system as an intermediary between the molding process and final quality control. This non-contact imaging and analysis system continuously monitors characters on molded parts, detecting molding flashing damage without interrupting production. The system serves as a mediator that enables continuous operation while maintaining quality control.
Solution Approach 2:
The patent replaces manual inspection methods with automated optical imaging and digital image analysis. Instead of mechanical or human operators physically examining each part (which would slow production), the system uses non-contact optical sensors and algorithmic analysis to continuously detect damaged identifiers, enabling high-speed continuous production while maintaining quality standards.
Data Source
AI summary
A circuit board processing system configured to process a panel, the system comprising a character recognition system configured to obtain a digital image of a character imprinted on a surface of the panel. The character recognition system is further configured to perform image analysis on the digital image of the character, the image analysis including applying a local threshold value to one or more localities of the digital image to provide a binarized image of the character; and to classify the character based on the image analysis, the local threshold value applied to each locality of the one or more localities being based on pixel information around each respective locality.