Character Recognition on Multiple Carriers via Segmented Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current character recognition methods are inefficient in recognizing characters on multiple carriers simultaneously due to high error rates caused by background interference and the inability to accurately distinguish characters on different carriers, leading to low recognition efficiency.
Innovation Solution
A character recognition method that involves feature extraction, candidate detection box screening, and angle adjustment to accurately identify and recognize characters on multiple carriers by using a neural network architecture with feature extraction, carrier and character detection networks, and an angle adjustment module to improve accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional character recognition methods are used on images with multiple carriers, then the recognition process is simple, but the recognition accuracy deteriorates due to background interference and inability to distinguish characters on different carriers
Solution Approach 1:
The patent applies segmentation by dividing the image processing into distinct stages: first extracting candidate carrier detection boxes to identify individual carriers, then extracting candidate character detection boxes within those carriers, and finally performing recognition. This multi-stage segmentation allows the system to handle multiple carriers separately, eliminating background interference from other carriers and improving recognition accuracy without overwhelming complexity
Solution Approach 2:
The patent implements preliminary action by performing carrier detection and candidate box screening before the actual character recognition. The system first identifies and outlines character carriers, screens candidate boxes to remove false positives, and prepares cleaned detection boxes for recognition. This preliminary processing eliminates background interference in advance, ensuring that only relevant character regions are processed during recognition
2Measurement precision
If character recognition is performed on multiple carriers simultaneously without screening, then the processing speed is fast, but the recognition accuracy deteriorates due to background interference from other carriers
Solution Approach 1:
The patent performs preliminary carrier detection and candidate box screening before recognition to eliminate background interference. By first identifying valid carriers and their associated character regions, the system ensures that only relevant regions are processed during recognition, improving accuracy without sacrificing overall efficiency
Solution Approach 2:
The patent segments the processing into carrier-level operations and character-level operations. By first detecting carriers and then processing characters within each carrier separately, the system avoids the computational burden of processing the entire image at once while maintaining high accuracy through focused processing of relevant regions
3Measurement precision
If all candidate detection boxes are processed for recognition, then the recognition completeness is high, but the processing time increases due to inclusion of false positive boxes from background
Solution Approach 1:
The patent extracts and removes false positive candidate boxes through screening operations. After initial candidate box generation, the system applies screening criteria to identify and eliminate boxes that do not correspond to actual characters on valid carriers. This extraction of irrelevant data before recognition reduces processing time while maintaining recognition completeness for valid characters
Solution Approach 2:
The patent performs preliminary screening of candidate boxes to remove false positives before the recognition stage. By validating carriers and filtering candidate boxes in advance, the system ensures that only high-probability character regions are processed during recognition, significantly reducing unnecessary processing time while maintaining high recognition accuracy
Data Source
AI summary
A character recognition method includes: performing feature extraction on an image to be recognized to obtain a first feature map; processing the first feature map to at least obtain N first candidate carrier detection boxes, each first candidate carrier detection box being configured to outline a region of a character carrier; screening the N first candidate carrier detection boxes to obtain K first target carrier detection boxes; performing a feature extraction on the first feature map to obtain a second feature map; processing the second feature map to obtain L first candidate character detection boxes, each first candidate character detection box being configured to outline a region containing at least one character; screening the L first candidate character detection boxes to obtain J first target character detection boxes; and recognizing characters in the J first target character detection boxes to obtain J target character informations.