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

Online object classification method and system based on fast similarity network fusion algorithm

A network fusion and target classification technology, which is applied in the field of online target classification methods and systems based on fast similarity network fusion algorithm, can solve problems such as uncollected data and no discovery, and achieve fast and efficient online target classification and good classification. performance, and the effect of improving classification accuracy

Active Publication Date: 2018-05-04
上海海维工业控制有限公司
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, there is no description or report of the similar technology of the present invention, and no similar data at home and abroad have been collected yet.

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
  • Online object classification method and system based on fast similarity network fusion algorithm
  • Online object classification method and system based on fast similarity network fusion algorithm
  • Online object classification method and system based on fast similarity network fusion algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0044] This embodiment provides an online target classification system based on a fast similarity network fusion algorithm, including: a video preprocessing module, a fast similarity network fusion analysis module, and a feedback labeling training module; wherein:

[0045]The video preprocessing module is used to extract the video foreground target as the initial training sample set, and extracts the image features of the target in the initial training sample set, and sends it to the fast similarity network fusion analysis module;

[0046] The fast similarity network fusion analysis module is used to fuse the image features of the target, analyze the similarity between the targets in the initial training sample set according to the fused similarity matrix, select samples that are difficult to classify, and send To the feedback annotation training module;

[0047] The feedback labeling training module is used to manually mark samples that are difficult to classify, and return t...

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 invention provides an online target classification system based on a fast similarity network fusion algorithm, comprising: a video preprocessing module for extracting video foreground targets as an initial training sample set, and extracting image features of targets in the initial training sample set, Send to the fast similarity network fusion analysis module; the fast similarity network fusion analysis module is used to fuse the image features of the target, and analyze the similarity between the targets in the initial training sample set according to the fused similarity matrix, Select the samples that are difficult to classify and send them to the feedback labeling training module; the feedback labeling training module is used to manually calibrate the samples that are difficult to classify, and return the manually calibrated samples to the video preprocessing module for the initial training sample set. renew. At the same time, the online target classification method of the above system is provided. The invention realizes fast and efficient online target classification.

Description

technical field [0001] The present invention relates to an online target classification method and system, in particular to an online target classification method and system based on a fast similarity network fusion algorithm. Background technique [0002] With the rapid development of social politics and economy, along with the occurrence of more and more social incidents and crimes, the attention paid to video surveillance is constantly increasing. The research on video target classification has great practical significance for the field of security monitoring. [0003] Among the existing object classification methods, some methods use shape and motion information for classification. This type of method is relatively fast, but its robustness is also limited because the shape of the object in the video will vary with the viewing angle. changes, the target classification accuracy is poor; another type of method considers the relationship between the calibrated data and a lar...

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): G06K9/66
Inventor 张重阳卢贤龙
Owner 上海海维工业控制有限公司
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