OCR Key Character Line Extraction for Recognition Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical Character Recognition (OCR) technologies face inefficiencies in processing large amounts of character information from images, leading to reduced recognition performance and user dissatisfaction due to the need to process non-key character lines, which can slow down the recognition process and reduce accuracy.
Innovation Solution
A method and apparatus that acquire images, analyze layouts to extract character area blocks, determine key character lines, and use OCR to recognize only the key character lines, allowing for user selection or automatic identification using keyword libraries to improve efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OCR technology processes all character information from images, then comprehensive recognition is achieved, but processing time increases and recognition accuracy decreases
Solution Approach 1:
The patent segments the image into multiple character area blocks based on layout analysis, then further segments these blocks into individual character lines. This hierarchical segmentation allows the system to process only relevant portions (key character lines) rather than all character information, thereby reducing processing time while maintaining recognition accuracy for important data.
Solution Approach 2:
The patent extracts and identifies key character lines from the segmented character area blocks based on predefined criteria or user selection. By extracting only the essential character lines that contain important information, the system avoids processing redundant data, thus reducing processing time while preserving recognition accuracy for critical information.
2Loss of information
If OCR technology processes all character lines in images, then complete information is captured, but processing efficiency decreases
Solution Approach 1:
The patent divides the image into character area blocks and further segments these into individual character lines. This segmentation enables selective processing where only key character lines are subjected to OCR recognition, while non-key lines are either skipped or processed with lower priority, thereby improving processing efficiency without completely sacrificing information completeness.
Solution Approach 2:
The patent applies partial action by performing full OCR processing only on identified key character lines rather than all character lines in the image. This selective approach achieves sufficient information capture for most applications while significantly improving processing efficiency by avoiding redundant processing of non-critical information.
3Measurement precision
If layout analysis is performed to extract character area blocks, then recognition precision is improved, but device complexity increases
Solution Approach 1:
The patent implements layout analysis that segments the image into character area blocks based on spatial distribution and visual characteristics. This segmentation approach improves recognition precision by organizing characters into meaningful groups while maintaining relatively simple processing logic through rule-based block identification and hierarchical structuring.
Data Source
AI summary
A method for information recognition using an Optical Character Recognition (OCR) program includes acquiring an image of an object to be recognized, analyzing a layout of the contents of the image and extracting character area blocks in the image, determining character lines in the character area blocks, and recognizing, by the OCR program, character information of the key character lines in the character area blocks.