Text Extraction from Images Using Edge Mask Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical Character Recognition (OCR) operations often fail to produce high-quality results when extracting text from natural or synthetic digital images due to the text being embedded within complex image content.
Innovation Solution
A method that generates an edges mask from the input image to identify probable text areas, performs OCR only on these areas, and overlays the extracted text onto a modified image to create an editable document, improving the accuracy and efficacy of text extraction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional OCR operations are applied to entire images with embedded text, then text extraction coverage is improved, but accuracy deteriorates due to complex image content interference
Solution Approach 1:
The patent divides the image processing task into distinct segments: first generating an edges mask to identify potential text regions, then performing OCR only on those segmented regions rather than the entire image. This segmentation approach isolates text areas from complex background content, improving accuracy while maintaining coverage
Solution Approach 2:
The patent applies different processing qualities to different regions of the image. Edges mask regions identified as potential text areas receive focused OCR processing, while non-text regions are excluded. This local quality approach concentrates computational resources on text-bearing areas, enhancing extraction accuracy without sacrificing coverage
2Loss of information
If OCR operations are performed on the entire input image, then comprehensive text extraction is achieved, but processing time increases due to unnecessary analysis of non-text areas
Solution Approach 1:
The patent performs preliminary action by generating an edges mask before executing OCR operations. This pre-processing step identifies and marks potential text regions, allowing the subsequent OCR process to focus only on relevant areas. This preliminary segmentation reduces processing time while maintaining complete text extraction
Solution Approach 2:
The patent applies partial action by performing OCR only on edges mask-identified text regions rather than the entire image. This selective approach avoids excessive processing of non-text areas, reducing computational time while ensuring all actual text regions are processed through the edges mask guidance
Data Source
AI summary
A method for extracting text from an input image and generating a document includes: generating an edges mask from the input image; generating an edges image that is derived from the edges mask; identifying, within the edges mask, one or more probable text areas; extracting a first set of text characters by performing a first optical character recognition (OCR) operation on each of one or more probable text portions, of the derived edges image, corresponding to each of the probable text areas; generating a modified image by erasing, from the input image, image characters corresponding to the first set of text characters extracted by the first OCR operation; and generating a document by overlaying the extracted first set of text characters on the modified image.