Multithreaded OCR System for Text Recognition Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional optical character recognition (OCR) systems face challenges in accurately detecting text from images with poor quality, such as variations in lighting, shadows, contrast, and blur, especially in real-world scenarios where ideal image capture is impractical.
Innovation Solution
The use of multiple OCR engines processing images in a multithreaded mode, with a confidencing module determining confidence scores based on various attributes, and a combination module generating a consensus string of text weighted by confidence scores from each engine, to improve text recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional OCR engines are used to process images, then processing speed is maintained, but text recognition accuracy deteriorates in low-quality images
Solution Approach 1:
The patent combines multiple OCR engines with different recognition algorithms and characteristics into a unified system. Each engine processes the input image independently, and their results are integrated through a combination module that weights outcomes based on confidence scores and engine performance, thereby improving overall accuracy while maintaining processing efficiency
2Measurement precision
If multiple OCR engines are used to improve accuracy, then text recognition precision improves, but system complexity increases
Solution Approach 1:
The patent introduces a combination module as an intermediary that manages the outputs from multiple OCR engines. This module receives recognition results from each engine, assigns confidence scores, and integrates the results through weighted combination, thereby simplifying the complexity of managing multiple engines while maintaining improved accuracy
3Measurement precision
If multiple OCR engines process images concurrently, then recognition accuracy improves, but processing time increases
Solution Approach 1:
The patent implements multithreaded concurrent processing where multiple OCR engines operate simultaneously on the input image rather than sequentially. This continuous parallel action allows the system to maintain high processing throughput while achieving improved accuracy through the integration of multiple recognition results
4Reliability
If conventional single-engine OCR is used, then system simplicity is maintained, but reliability deteriorates in varying image conditions
Solution Approach 1:
The patent changes the operational parameters of the OCR system by employing multiple engines with different recognition algorithms and characteristics. The combination module dynamically adjusts weighting parameters based on confidence scores and engine performance, enabling the system to adapt to varying image conditions and maintain high reliability across different quality levels
Data Source
AI summary
Various embodiments enable a device to perform tasks such as processing an image to recognize and locate text in the image, and providing the recognized text an application executing on the device for performing a function (e.g., calling a number, opening an internet browser, etc.) associated with the recognized text. In at least one embodiment, processing the image includes substantially simultaneously or concurrently processing the image with at least two recognition engines, such as at least two optical character recognition (OCR) engines, running in a multithreaded mode. In at least one embodiment, the recognition engines can be tuned so that their respective processing speeds are roughly the same. Utilizing multiple recognition engines enables processing latency to be close to that of using only one recognition engine.