Text Recognition Module for Unknown Script Orientation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical character recognition (OCR) systems face challenges in identifying the script and orientation of text in images, especially when the language is unknown, as they rely on prior knowledge of character shapes and patterns that are language-specific, making it difficult to determine the script and language without prior information.
Innovation Solution
A computer-implemented method and system that trains a text recognition module by receiving images with known text, identifying connected components, generating word fragments, and using these to train the module to recognize text in unknown scripts and orientations, allowing for the determination of script and orientation in images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OCR systems use prior knowledge of character shapes and patterns, then recognition accuracy improves, but the system cannot handle unknown scripts and languages
Solution Approach 1:
The system performs preliminary action by training the text recognition module in advance using a diverse corpus of text images in multiple scripts. This pre-training enables the module to recognize character shapes and patterns across different scripts before actual OCR processing, allowing it to handle unknown scripts without requiring prior knowledge specification.
Solution Approach 2:
The text recognition module is designed with universality by implementing multi-script support through training on diverse language corpora. The module can recognize characters from multiple scripts (Latin, Cyrillic, Greek, Arabic, Hebrew, etc.) using a single unified model, eliminating the need for separate language-specific configurations and enabling automatic adaptation to unknown scripts.
2Reliability
If the system requires prior knowledge of language and script, then OCR processing is more accurate, but the system cannot process images with unknown languages
Solution Approach 1:
The system implements self-service by automatically detecting the script and language of input text images without requiring user specification. The text recognition module autonomously adapts to the script present in the image through its pre-trained multi-script capabilities, and the system automatically selects appropriate processing parameters, eliminating the need for manual language configuration and making the system easy to operate with unknown languages.
3Measurement precision
If multiple language-specific OCR modules are created, then each language achieves high accuracy, but system complexity increases
Solution Approach 1:
The system merges multiple language-specific OCR capabilities into a single unified text recognition module. By training one module on a comprehensive corpus covering multiple scripts and languages, the system consolidates what would otherwise require separate modules into one integrated solution, reducing system complexity while maintaining high recognition accuracy across all supported languages through shared feature extraction and recognition mechanisms.
Data Source
AI summary
A page layout module receives a page image displaying text in an unknown script and unknown orientation, determines a text section in the received image and transmits the text section to an orientation and script module. The orientation and script module comprises a training module, a classifier and a recognition module. The training module trains the classifier to identify connected components that include a connected portion of one or more characters of text. The recognition module uses the trained classifier to identify in the received text section a set of connected components. The recognition determines the likely orientation and script for the connected components and then uses the determined information to determine the orientation and script for the text section. The determined orientation and script for the text section is transmitted to the OCR module. The OCR module uses the determined orientation and script to recognize text in the text section.