Image-Text Matching in Specifications Using Preference Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Quality control personnel face time-consuming and error-prone manual processes in matching illustrations and text descriptions in specifications, lacking an efficient method for automating this process.
Innovation Solution
A matching system utilizing an image-and-text recognition device, preference-value calculation device, filtering device, and recommendation device to automate the matching of image blocks and text blocks in specifications, calculating preference values based on position and content characteristics to improve efficiency and reduce errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual marking and matching is used for illustrations and text descriptions in specifications, then flexibility and adaptability are maintained, but time consumption increases and error rates rise
Solution Approach 1:
The patent replaces the manual mechanical marking process with an automated image recognition and text extraction system. The system uses computer vision to detect illustrations and optical character recognition (OCR) to extract text descriptions, automatically matching them without human intervention. This substitution of manual mechanical operations with automated digital processing directly reduces time consumption while maintaining or improving matching accuracy through algorithmic precision.
Solution Approach 2:
The system enables self-service by automatically processing specifications without requiring quality control personnel to manually mark items. The automated system independently performs image recognition, text extraction, and matching operations, allowing the specification document to be processed autonomously. This self-service capability eliminates the time-consuming manual marking process while maintaining reliable matching through automated algorithms.
2Productivity
If automated image and text recognition is implemented, then marking efficiency improves, but system complexity increases
Solution Approach 1:
The patent segments the complex automated matching process into distinct functional modules: image recognition module for detecting illustrations, text extraction module for OCR processing, preference value calculation module for scoring based on position and content, and matching module for final pairing. This segmentation divides the complex system into manageable, independent components that can be developed, tested, and maintained separately, reducing the perceived complexity while maintaining high productivity.
Solution Approach 2:
The system introduces an intermediary preference value calculation mechanism that mediates between image recognition and text extraction results. This intermediary layer calculates preference values based on positional relationships and content characteristics, serving as a bridge that harmonizes different recognition results before final matching. This intermediary step simplifies the overall system architecture by providing a standardized scoring mechanism that reconciles multiple data sources.
3Manufacturing precision
If manual marking is performed based on experience, then adaptability to different specifications is maintained, but consistency and precision decrease
Solution Approach 1:
The patent transforms the subjective experience-based marking process into objective parameter-based automated matching. Instead of relying on operator experience and judgment, the system uses measurable parameters such as positional distance between illustrations and text, content similarity scores, and spatial relationship metrics. These quantifiable parameters enable precise, consistent matching that is reproducible and independent of human operator variability, directly improving marking precision while maintaining ease of operation through automated processing.
Data Source
AI summary
Provided is a matching system for images and text descriptions in a specification. The matching system includes an image-and-text recognition device, receiving a specification and recognizing image blocks and text blocks thereon, the image block having corresponding covering range; and a preference value calculation device, assigning preference value to each of the text blocks according to positional relationship between the above-mentioned text block and the above-mentioned image block, and the contents of the above-mentioned text block, for matching the image blocks and the text blocks.