OCR Text Completion via Segmented Sub-Image Radius Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Images containing partially covered text strings pose challenges in summarization and translation, leading to inaccurate results due to incomplete text extraction.
Innovation Solution
A method utilizing an Optical Character Reader (OCR) to extract text from images, analyzing context, and performing image searches within segmented sub-images to identify and complete hidden text strings across multiple radius levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If OCR is used to extract text from images, then text extraction speed is improved, but text extraction accuracy deteriorates when text is partially covered
Solution Approach 1:
The patent segments the image into multiple sub-images around the detected text region at different radius levels. By dividing the search space into concentric circular regions, the system can efficiently search for covering objects at multiple scales without processing the entire image, thus maintaining speed while improving accuracy in identifying partially covered text.
Solution Approach 2:
The patent introduces a radial dimension by searching at multiple radius levels around the text region. Instead of only checking the immediate text area, the system expands the search in concentric circles, adding a dimensional aspect to the text completion process that enables finding covering objects that may be partially obscuring the text.
2Measurement precision
If image search is performed to identify complete text strings, then text extraction accuracy is improved, but processing time increases
Solution Approach 1:
The patent divides the image search into segmented sub-images at different radius levels around the detected text. This segmentation allows the system to focus computational resources on specific regions likely to contain covering objects, rather than searching the entire image, thus reducing processing time while maintaining accuracy.
Solution Approach 2:
The patent performs partial image search by only examining regions at specific radius levels around the text, rather than conducting a complete full-image search. This partial action approach is sufficient to find covering objects while significantly reducing the computational burden and processing time.
3Measurement precision
If multiple radius levels are searched to find complete text, then text completion accuracy is improved, but system complexity increases
Solution Approach 1:
The patent uses segmentation to divide the search space into manageable sub-images at different radius levels. This structured division simplifies the overall system by breaking down the complex task of finding covering objects into smaller, more manageable regional searches, making the multi-level search process more tractable.
Solution Approach 2:
The patent performs preliminary detection of text regions before conducting the multi-level radius search for covering objects. This preliminary action identifies the center point and initial text content, which then guides the subsequent radius-level searches, reducing the overall system complexity by establishing a clear sequential workflow.
Data Source
AI summary
Provided are techniques for image text analysis for identifying hidden text. An Optical Character Reader (OCR) is utilized to extract a text string from an image. Context within the image is analyzed. It is determined that the extracted text string is a partial text string based on the context. For a first radius level of a plurality of radius levels, a segmented sub-image is identified around the partial text string within the first radius level, an image search on the segmented sub-image is performed to identify a candidate text string, and, in response to determining that the candidate text string is a complete text string, the complete text string is provided for performing an action.