OCR Image Preprocessing via ROI Feature Descriptor Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical character recognition (OCR) systems face challenges in efficiently processing images with low contrast, noise, character variation, and non-ideal factors, leading to segmentation errors and unnecessary processing of non-relevant text, which affects throughput and computational resources, especially in high-speed manufacturing scenarios.
Innovation Solution
An image processing system that autonomously detects and isolates regions of interest (ROI) within captured images, transforming them into feature descriptor representations for comparison with templates, allowing the OCR engine to focus on relevant text while ignoring non-ROI content, thereby optimizing processing and reducing computational burden.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If OCR processing is performed on the entire captured image including non-ROI content, then all text in the image can be recognized, but computational resources are wasted on needless processing and throughput is reduced
Solution Approach 1:
The patent divides the captured image into multiple regions of interest (ROIs) using feature descriptor representations and template matching. By segmenting the image into distinct ROIs, the system processes only relevant text areas, eliminating wasted computational resources on non-ROI content while maintaining recognition accuracy for all necessary text.
Solution Approach 2:
The patent extracts and processes only the regions of interest from the captured image by comparing feature descriptor representations against templates. This extraction approach isolates relevant text areas from the rest of the image, allowing the OCR engine to focus computational resources exclusively on meaningful content and improve overall processing throughput.
2Measurement precision
If image preprocessing is performed to handle low contrast and noise, then character recognition accuracy improves, but processing time and computational complexity increase
Solution Approach 1:
The patent performs preliminary preprocessing operations including transforming the captured image into a feature descriptor representation and comparing it against templates before full OCR processing. This preliminary action identifies and isolates ROIs in advance, allowing subsequent OCR processing to focus only on relevant areas, thereby improving recognition accuracy while minimizing overall processing time.
3Reliability
If the OCR engine processes all text in the captured image, then comprehensive text recognition is achieved, but processing speed decreases due to unnecessary computation on non-relevant content
Solution Approach 1:
The patent applies different processing qualities to different regions of the image by identifying and isolating regions of interest through template matching and feature descriptor comparison. High-quality OCR processing is applied only to identified ROIs containing relevant text, while non-ROI areas receive minimal or no processing, thereby maintaining recognition completeness for necessary text while significantly improving processing speed.
Data Source
AI summary
A captured image contains a region of interest (ROI) including a plurality of characters to be recognized as text, and non-ROI content to be excluded from the OCR. The captured image is preprocessed to detect and locate the ROI in the captured image, and to determine a boundary of the ROI, including transforming the captured image to a first feature descriptor representation (FDR), and performing a comparison between the first FDR and at least one ROI template that includes at least a second FDR of a representative ROI image. The preprocessing produces an output to be provided to an OCR engine to perform autonomous OCR processing of the ROI while ignoring the non-ROI content based on the determined boundary of the ROI.