Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A method and device for image retrieval

A technology of pictures and target pictures, which is applied in still image data retrieval, metadata still image retrieval, character and pattern recognition, etc., can solve the problem of not being able to search for target pictures, and achieve the effect of alleviating recognition errors

Active Publication Date: 2020-07-03
BEIJING YUNCE INFORMATION TECH CO LTD
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The embodiment of the present application provides a method and device for image retrieval to alleviate the problem in the prior art that the target image cannot be searched due to the wrong recognition of the text in the image

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A method and device for image retrieval
  • A method and device for image retrieval
  • A method and device for image retrieval

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] This application provides a picture retrieval method to alleviate the problem in the prior art that the target picture cannot be searched due to the wrong recognition of the text in the picture, see figure 1 , this method includes:

[0035] Step 11: Perform text recognition on the image to be retrieved, and determine the character string to be retrieved corresponding to the image to be retrieved according to the result of the text recognition.

[0036] Specifically, to perform text recognition on the image to be retrieved, OCR technology or other text recognition technologies may be used. OCR technology determines the shape of text in a picture by detecting dark and light patterns, and then uses character recognition methods to translate the shape into computer text. For the characters in the picture, the text in the image is converted into a text format by the recognition software, which can be further edited and processed by the word processing software.

[0037] Am...

Embodiment 2

[0119]Based on the solutions provided in the above embodiments, this embodiment provides a method for image retrieval. In this embodiment, the target character string is "one-click application", and the text in the image to be retrieved is "one-key application". Recognition error, "key" is recognized as "chain", so that the character string to be retrieved is "a chain application". In this case, the implementation process of the method provided by this application is as follows:

[0120] The target four-corner number corresponding to the target string "one-click application" is "10000857405000635727", and the four-corner number to be retrieved corresponding to the search string "one-chain application" is "10000847355000635727". The above method of converting a string to a four-corner number is implemented by reference The method described in Example 1 is converted according to the four-corner number rule. In this example, the four-corner number corresponding to one character is...

Embodiment 3

[0142] Based on the solutions provided by the above embodiments, the embodiments of the present application also provide a picture retrieval device, such as Figure 7 shown, including:

[0143] Character string determining module 71, carry out text recognition to the image to be retrieved, and determine the character string to be retrieved corresponding to the image to be retrieved according to the result of character recognition;

[0144] The four-corner number generation module 72 generates the four-corner number to be retrieved according to the character string to be retrieved;

[0145] The similarity determining module 73 calculates the similarity between the four-corner number to be retrieved and the target four-corner number;

[0146] The target picture determining module 74 determines the picture to be retrieved as the target picture corresponding to the target four-corner number when the similarity satisfies a preset condition.

[0147] This application converts the ...

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

No PUM Login to View More

Abstract

The present application relates to the field of terminals, and in particular to a method and device for image retrieval, which are used to alleviate the problem in the prior art that a target image cannot be searched due to misrecognition of text in an image. The method comprises: performing character recognition on the image to be retrieved, determining a character string to be retrieved corresponding to the image to be retrieved according to the result of character recognition; generating a four-corner number to be retrieved according to the character string to be retrieved; calculating the ratio between the four-corner number to be retrieved and the target four-corner number Similarity: when the similarity satisfies the preset condition, the picture to be retrieved is determined as the target picture corresponding to the target four-corner number. This solution judges whether the searched string matches the text in the picture based on the four-corner number rule. When there is a font difference between the text recognized from the picture and the original text in the picture, this application compares the searched string with the recognized text based on the four-corner number. string, which can alleviate to a certain extent the situation that the target image cannot be searched due to recognition errors.

Description

technical field [0001] The present application relates to the field of terminals, in particular to a method and device for image retrieval. Background technique [0002] When people search for images online, they often do so by text. For pictures containing text, in the prior art, the text in the picture is often converted into text form through Optical Character Recognition (OCR). When the user's search keyword matches the text recognized from the picture , the image can be regarded as the target image searched by the user. [0003] However, the fonts of the text in the picture are various, and some of the text has been designed and beautified, which is quite different from the original font. For the picture with low pixels, the small text in the picture is not clear and incomplete. OCR is suitable for the above The recognition accuracy of the text in the picture is low, which leads to the user not being able to search for the target picture. Contents of the invention ...

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
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/58G06K9/34
CPCG06V30/153
Inventor 戴亦斌陈雪军
Owner BEIJING YUNCE INFORMATION TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com