Text Detection in Digital Engineering Drawings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current text detection methods, such as stroke width transform, fail to differentiate between text and non-text regions in digital engineering drawings due to both having constant stroke widths, rendering them unreliable for automatic knowledge extraction.
Innovation Solution
A method and system that convert digital engineering drawings to binary images, detect lines inclined at angles, remove non-text regions based on height, width, and pixel density, and process text regions using connected component analysis and OCR techniques to identify and extract text.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If stroke width transform is used for text detection, then text detection can be performed on general images, but it fails to differentiate text from non-text regions in engineering drawings where both have constant stroke width
Solution Approach 1:
The patent applies local quality by analyzing specific local characteristics of regions in engineering drawings. Instead of using a general stroke width transform, the method examines local pixel density, connectivity patterns, and geometric properties within specific regions to differentiate text from non-text elements, both of which have constant stroke width. This localized analysis enables reliable text detection in the specific context of engineering drawings.
2Device complexity
If general text detection methods are applied to engineering drawings, then processing can be simplified, but the detection reliability deteriorates due to inability to distinguish text from non-text components
Solution Approach 1:
The patent changes the parameters used for text detection from general stroke width measurements to specific parameters suitable for engineering drawings. The method uses pixel density thresholds, connectivity analysis, and geometric property measurements that are tailored to the constant-stroke-width characteristics of engineering drawing elements. These parameter changes enable reliable differentiation between text and non-text regions while maintaining manageable processing complexity.
3Adaptability or versatility
If stroke width transform is used, then the detection process can be applied broadly, but it cannot differentiate between text and non-text regions in digital engineering drawings
Solution Approach 1:
The patent applies segmentation by dividing the engineering drawing into multiple regions and analyzing each region's specific properties. The method segments the image into candidate text regions and non-text regions, then applies different analysis criteria to each segment. This includes examining pixel density distributions, connectivity patterns, and geometric properties within each segmented region to differentiate text from non-text elements effectively.
Data Source
AI summary
The present disclosure discloses method and text detection system for detecting text in digital engineering drawings. The text detection system receives digital engineering drawing, where digital image is converted to binary image. One or more lines inclined at one or more angles are detected based on pre-defined parameters, where one or more lines are removed from binary image. The text detection system determines image contours in binary image based on connected pixels with similar characteristics and identify one or more regions in binary image, to be non-text regions based on height, width, and pixel density, of corresponding image contours and predefined minimum height, minimum width, maximum height, maximum width, and predefined pixel density, where non-text regions are removed from binary image. The text detection system determines text regions in binary image based on removal of non-text regions and processing text regions to detect text in binary image of engineering drawing.