Character Recognition with Confidence-Guided Meta-Information Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional character string recognition systems face reduced accuracy when meta-information is incorrect or mismatched with the input image, leading to incorrect recognition results.
Innovation Solution
An information processing apparatus that utilizes a directed acyclic graph structure to define the order of meta-information, allowing for flexible updating and correction of meta-information based on confidence scores and update histories, thereby improving recognition accuracy by iteratively refining the meta-information used in character recognition processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If meta-information is used to improve recognition accuracy, then recognition accuracy is improved, but the system becomes vulnerable to incorrect or mismatched meta-information
Solution Approach 1:
The patent implements a feedback mechanism where the character recognition unit provides feedback on the reliability of recognition results to the meta-information management unit. This feedback loop allows the system to detect when meta-information is incorrect or mismatched and automatically adjust its behavior, thereby maintaining reliability while preserving the accuracy benefits of using meta-information.
Solution Approach 2:
The patent makes the meta-information management dynamic by allowing the meta-information management unit to switch between different character recognition units based on the reliability of feedback. Instead of using a fixed meta-information set, the system dynamically adapts its meta-information based on real-time performance feedback, resolving the contradiction between using meta-information for accuracy and maintaining system reliability.
2Measurement precision
If different recognition processes are performed for each field type, then recognition accuracy is improved, but device complexity increases
Solution Approach 1:
The patent implements a universal character recognition unit that can handle multiple field types through a single unified process. Instead of requiring separate recognition processes for each field type, the system uses one multi-functional recognition unit that adapts to different meta-information types, thereby reducing device complexity while maintaining the accuracy benefits of field-type-specific processing.
Solution Approach 2:
The patent changes the approach from structural complexity (multiple separate processes) to parameter-based differentiation (using meta-information parameters to guide a single process). By using parameter changes in the recognition process based on meta-information rather than creating separate processes for each field type, the system achieves field-type-specific accuracy without the complexity of multiple dedicated processes.
Data Source
AI summary
An information processing apparatus includes a memory and one or more hardware processors. The memory stores order information in which an order of pieces of meta-information for a character to be recognized is defined. The one or more hardware processors are connected to the memory and function as a recognition unit and an update unit. The recognition unit serves to perform character recognition on an image including a character string by using first meta-information specified from the pieces of the meta-information. The update unit serves to update the first meta-information to second meta-information in accordance with the order information in a case when a confidence score of the character recognition satisfies a predetermined condition. The character recognition is performed by using the second meta- information.