Text Recognition via Edge-Based Image Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical character recognition (OCR) processes often produce inaccurate text from blurry or skewed images, especially when non-standard text and objects are present, requiring significant human and computing resources for correction and processing.
Innovation Solution
A text recognition platform uses computer vision to identify edges in image data, crops segments of interest to exclude edges that hinder OCR, and performs OCR on these segments to improve accuracy, enabling verification and formatting of recognized text.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OCR is performed on the entire image data including edges, then complete text coverage is achieved, but OCR accuracy deteriorates due to distracting edges and non-standard objects
Solution Approach 1:
The patent divides the image data into multiple segments based on detected edges, identifying a segment of interest that contains the text while excluding distracting edges and non-standard objects. This segmentation allows OCR to be applied selectively to the relevant portion, improving accuracy while maintaining text coverage.
Solution Approach 2:
The patent extracts the segment of interest from the full image data by identifying and isolating the region containing text based on edge analysis. This extraction removes distracting edges and non-standard objects from the OCR processing area, thereby improving recognition accuracy without losing the target text information.
2Productivity
If OCR is performed on the entire image data, then all text is processed, but resource consumption increases significantly
Solution Approach 1:
The patent segments the image data to identify only the relevant segment containing text, excluding irrelevant areas with edges and non-standard objects. This segmentation enables OCR to process only the necessary portion of the image, reducing computing resource consumption while maintaining processing throughput.
Solution Approach 2:
The patent applies partial action by performing OCR only on the identified segment of interest rather than the entire image data. This selective processing reduces the computational burden and energy consumption while still achieving complete text extraction from the relevant areas.
3Measurement precision
If edges are included in the image data for OCR, then complete image information is preserved, but text recognition accuracy deteriorates
Solution Approach 1:
The patent extracts and removes distracting edges from the image data by identifying them through edge detection algorithms. The segment of interest containing text is separated from the edges, allowing accurate text recognition while the edges are excluded from the OCR processing area.
Solution Approach 2:
The patent segments the image data into distinct regions based on edge detection, separating the text-containing segment from the edge-containing segments. This segmentation enables the system to preserve complete image information for analysis while excluding only the distracting edges from the actual OCR processing.
Data Source
AI summary
A device may receive image data representing a document, the document including: text, and edges. Based on the edges, the device may identify, a segment of interest within the image data and crop the segment of interest to obtain a portion of the image data. In addition, the device may perform optical character recognition on the portion of the image data, the optical character recognition producing recognized text. The device may obtain, based on the recognized text, validation data that includes verification text, and determine whether the recognized text is verified based on the verification text. Based on a result of the determination, the device may perform an action.