Image Processing Masking for Object Extraction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods struggle to accurately extract objects from images when symbols and other objects are misaligned or closely printed, leading to erroneous extraction during inspection.
Innovation Solution
An image processing method and apparatus that includes a symbol reader to set a mask area containing the symbol, and an object recognition component to recognize objects outside this area, improving extraction accuracy by masking the symbol area and using statistical scores to convert pixel components for accurate object recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a predetermined area of the image is set as a common data area and another predetermined area is set as a character data area for inspection, then the efficiency of inspection is improved, but the extraction accuracy of objects is deteriorated when symbols and objects are misaligned or closely printed
Solution Approach 1:
The image is divided into multiple inspection areas including a common data area, a character data area, and a symbol area. Each area is processed separately with appropriate reference patterns, allowing the system to maintain high inspection efficiency while improving object extraction accuracy by preventing misalignment issues between different data types.
Solution Approach 2:
The symbol area is extracted and processed separately from the common data area and character data area. By taking out the symbol area for independent processing with symbol-specific reference patterns, the system avoids erroneous extraction of objects when symbols and characters are closely printed or misaligned.
2Ease of operation
If multiple objects are printed close to one another to improve user convenience, then the usability is improved, but the extraction accuracy of objects is deteriorated due to misalignment during transportation
Solution Approach 1:
The inspection area is segmented into distinct zones including symbol area, common data area, and character data area. Each zone is processed with dedicated reference patterns, enabling accurate object extraction even when multiple objects are printed close together and may become misaligned during transportation.
Solution Approach 2:
Different reference patterns are applied to different areas of the image. The symbol area uses symbol-specific reference patterns, while the common data and character areas use their respective reference patterns. This local quality approach ensures accurate object extraction in each region regardless of overall misalignment.
Data Source
AI summary
An image processing method improves the extraction accuracy of objects other than symbols. The image processing method may include four processes. In the first process, an image is input. In the second process, a symbol in the image is read. In the third process, a mask area including the symbol is set. In the fourth process, an object located in an area other than the mark area in the image is recognized.