Methods and systems for performing on-device image to text conversion
a technology of image and text, applied in the field of performing image to text conversion, can solve the problems of generating errors in downstream tasks such as visual question answering (vqa), existing ocr solutions have no understanding of user edited text, and cannot solve problems such as highlighting text, and converting complex text of images
Pending Publication Date: 2022-10-27
SAMSUNG ELECTRONICS CO LTD
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Benefits of technology
The technical effect of this patent is to provide methods and systems for performing on-device image to text conversion, which aims to address the problems and disadvantages of existing methods and improve their efficiency and accuracy.
Problems solved by technology
However, existing OCR solutions have no understanding of user edited text like highlighted text, strikethrough, insert, and the like.
Thus, resulting in errors in downstream tasks like Visual Question Answering (VQA).
Also, the existing OCR solutions may produce errors while converting complex text of the image even though the text is present elsewhere in other clear regions of the image.
In addition, language selection from the image is a drawback in many Natural Language Processing (NLP) and vision tasks, since a default language may be taken as a device locale even if the image is in different language.
However, the ML Kit supports a Latin language / script as default and does not support other scripts.
Thus, the cloud based OCR solutions are neither scalable to devices due to huge memory usage and power consumption nor respect a user privacy since the image has to be uploaded to a server.
Also, the script based OCR has lesser accuracy than the language based OCR.
In such a scenario, the converted text may include errors since the existing OCR solutions consider the device locale as the default language.
In such a scenario, the converted text may include error with respect to the text in the complex font.
However, the existing OCR solutions do not use the text in clear and simple fonts for correcting the error in the text of complex fonts since the existing OCR solutions consider a dictionary of words or global knowledge to correct the extracted text from the image.
However, the existing OCR solutions may ignore such user edited portions while converting the image into the text since the existing OCR solutions have no understanding of the user edited document images.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreExamples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0057]The following description with reference to accompanying drawings provided to assist in a comprehensive understanding of various embodiments of the disclosure as defined by the claims and their equivalents. It includes various specific details to assist in that understanding but these are to be regarded as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the various embodiments described herein can be made without departing from the scope and spirit of the disclosure. In addition, descriptions of well-known functions and constructions may be omitted for clarity and conciseness.
[0058]The terms and words used in the following description and claims are not limited to the bibliographical meanings, but, are merely used by the inventor to enable a clear and consistent understanding of the disclosure. Accordingly, it should be apparent to those skilled in the art that the following description of various embod...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more PUM
Login to view more Abstract
A method and system for performing on-device image to text conversion are provided. Embodiments herein relates to the field of performing image to text conversion and more particularly to performing on-device image to text conversion with an improved accuracy. A method performing on-device image to text conversion is provided. The method includes language detection from an image, understanding of text in an edited image and using a contextual and localized lexicon set for post optical character recognition (OCR) correction.
Description
CROSS-REFERENCE TO RELATED APPLICATION(S)[0001]This application is a continuation application, claiming priority under § 365(c), of an International application No. PCT / KR2022 / 002031, filed on Feb. 10, 2022, which is based on and claims the benefit of an Indian Provisional Specification patent application number 202141005677, filed on Feb. 10, 2021, in the Indian Intellectual Property Office, and of an Indian Complete Specification patent application number 202141005677, filed on Feb. 3, 2022, in the Indian Intellectual Property Office, the disclosure of each of which is incorporated by reference herein in its entirety.TECHNICAL FIELD[0002]The disclosure relates to the field of performing image to text conversion. More particularly, the disclosure relates to performing on-device image to text conversion with an improved accuracy.BACKGROUND ART[0003]Optical Character Recognition (OCR) is an electronic or mechanical conversion of images into machine-readable form / text, which has to be...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more Application Information
Patent Timeline
Login to view more Patent Type & Authority Applications(United States)
IPC IPC(8): G06V30/148G06V30/14G06V30/22
CPCG06V30/153G06V30/1444G06V30/22G06V10/82G06V30/148G06V30/2455G06V30/226G06V30/147G06V30/18076G06V30/268
Inventor MOHARANA, SUKUMARRAMENA, GOPIMUNJAL, RACHIT SGOYAL, MANOJMOHARIR, RUTIKAARORA, NIKHILPRABHU, ARUN DVATSAL, SHUBHAM
Owner SAMSUNG ELECTRONICS CO LTD
Who we serve
- R&D Engineer
- R&D Manager
- IP Professional
Why Eureka
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Social media
Try Eureka
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap