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

Method and device for screening image data of character clothing

An image data and screening method technology, applied in the field of image recognition, can solve the problems of low crawling efficiency, inaccurate data acquisition, and inconvenient acquisition of design data.

Active Publication Date: 2018-12-21
SHANGHAI YUEPU INVESTMENT CENT (LLP)
View PDF6 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In the existing technology, since only pictures from the whole site or designated locations can be crawled, there are a large number of pictures that do not meet the requirements after crawling, such as pictures of advertisements, pictures of operation buttons, and low-quality thumbnails And so on, resulting in low crawling efficiency, inaccurate data acquisition, and inconvenience for further image recognition and design data acquisition

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
  • Method and device for screening image data of character clothing
  • Method and device for screening image data of character clothing
  • Method and device for screening image data of character clothing

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0057] refer to figure 2 , the first embodiment of the present invention provides a method for screening image data of people's clothing, including:

[0058] Step S10, acquiring a large number of images to be screened from preset websites;

[0059] As mentioned above, the method provided in this embodiment is based on computer-based image recognition technology to recognize a large number of images to be screened that include clothing. Among them, the massive images to be screened can be display pictures of models wearing clothes, catwalk pictures, life pictures, or pure clothing pictures without models.

[0060] As mentioned above, in this embodiment, the source of the target image can be aimed at a large number of photos of people wearing different styles of clothing or photos of simple clothing on clothing websites of different brands, and clean the image data before identifying the above photos , in order to reduce noise, remove useless, unrecognizable pictures, thereby...

Embodiment 2

[0075] refer to image 3 , the second embodiment of the present invention provides a method for screening character clothing image data, based on the above figure 2 In the first embodiment shown, after the step S30, after "screening out a large number of images to be screened whose image specification data reaches the preset image specification, as coarse-screen specification images", it also includes:

[0076] Step S50, obtaining the column list of the massive images to be screened corresponding to the coarse-screened image in the preset website, and obtaining column classification information corresponding to the coarse-screened image;

[0077] As mentioned above, the default website includes many column categories, such as: men's clothing, women's clothing, children's clothing, handbags, men's shoes, women's shoes, etc., and the website contains a large number of coarse-screened specification images. When the image is acquired, a large number of images of coarse sieve spe...

Embodiment 3

[0095] refer to Figure 4 , the third embodiment of the present invention provides a method for screening character clothing image data, based on the above image 3 In the second embodiment shown, the step S10, after "obtaining a large number of images to be screened from preset websites", further includes:

[0096] Step S120, acquiring EXIF ​​information of each of the massive images to be screened; constructing a current time node archive according to the image specification data and the EXIF ​​information corresponding to each of the massive images to be screened;

[0097] As mentioned above, the time node archive is an archive that characterizes the features corresponding to all the massive images to be screened. Wherein, it includes the image specification data and the EXIF ​​information of all currently obtained massive images to be screened.

[0098] As mentioned above, it should be noted that EXIF ​​information is an abbreviation of Exchangeable Image File, which is ...

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 invention provides a method and a device for screening the image data of a person's clothing, wherein the method comprises the following steps: obtaining a mass image to be screened of a preset website; Extracting image specification data; screening out the image of coarse sieve specification; an image recognition being performed on each of the rough screen specification images, and a character dress screening image being screened out. By cleaning the pictures obtained by the preset website, First of all, the images that meet the preset image specifications are filtered out, and then screens out the pictures containing the characteristics of the characters' costumes, At last, that character dress screen image is obtained, the data clean is completed, a large number of useless picturesare reduce, the data cleaning efficiency is high, and finally the image data after cleaning is accurate, which provides convenience for the recognition of the cleaned image and improves the work efficiency of the further image recognition to a certain extent.

Description

technical field [0001] The present invention relates to the technical field of image recognition, and more specifically, to a method and device for screening image data of clothing of characters. Background technique [0002] A website (Website) refers to a collection of web pages for displaying specific content made on the Internet according to certain rules and using tools such as HTML (an application under the standard general markup language). Different websites have different content and pictures. [0003] Clothing brand websites contain a large number of pictures of different styles of the current season and past seasons, including photos of models and pictures of only clothing without models. If you need to master the clothing styles of this brand or multiple brands And acquisition, you can crawl the pictures in the brand's related sites, and then identify the model's clothing collocation or clothing size, color, specification, style, dressing style and other informa...

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 Applications(China)
IPC IPC(8): G06K9/62
CPCG06F18/2113G06F18/22G06F18/214
Inventor 马修·罗伯特·斯科特郭胜王海涵唐颖雯
Owner SHANGHAI YUEPU INVESTMENT CENT (LLP)
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