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

Bilinear pedestrian re-identification network construction method based on body structure division

A network construction and pedestrian recognition technology, which is applied in the field of computer vision pedestrian re-identification, can solve the problem of low accuracy of pedestrian re-identification, achieve high discrimination and robustness, improve the degree of fusion, and improve the matching accuracy.

Active Publication Date: 2019-04-12
ACADEMY OF BROADCASTING SCI STATE ADMINISTATION OF PRESS PUBLICATION RADIO FILM & TELEVISION +1
View PDF4 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art, propose a bilinear pedestrian re-identification network construction method based on body structure division, and solve the problem of low pedestrian re-identification accuracy

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
  • Bilinear pedestrian re-identification network construction method based on body structure division
  • Bilinear pedestrian re-identification network construction method based on body structure division
  • Bilinear pedestrian re-identification network construction method based on body structure division

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] Embodiments of the present invention will be described in further detail below in conjunction with the accompanying drawings.

[0037] A bilinear person re-identification network based on body structure division, such as figure 1 shown, including the following steps:

[0038] Step 1. Construct the structural frame prediction sub-network, divide the original pedestrian image into body structure blocks, obtain six structural sub-frames, and combine these six sub-frames into a new pedestrian image.

[0039] Such as figure 2 As shown, the specific implementation method of this step is as follows:

[0040] ⑴ Key point positioning: Locate 14 key points of the body and use them as the basis for the division of structural blocks. point), left leg (3 points up, middle and down), right leg (3 points up, middle and down);

[0041] (2) Structural block prediction: According to 14 key points, use six rectangular structural frames to frame the corresponding body parts: head, upp...

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 relates to a bilinear pedestrian re-identification network construction method based on body structure division, and the method comprises the following steps: carrying out the body structure partitioning of an original pedestrian image, obtaining a plurality of structure sub-frames, enabling the plurality of sub-frames to be combined into a new pedestrian image, and constructing a structure frame prediction sub-network; setting a weighted local loss function to train the structure frame prediction sub-network; constructing two sub-networks, respectively taking the original pedestrian image and the recombined pedestrian image as input, and correspondingly extracting a global pedestrian feature and a local pedestrian feature; setting a bilinear fusion layer, and taking the bilinear fusion layer as a fusion layer of global features and local features to obtain final pedestrian feature representation; and training the whole network to obtain a bilinear pedestrian re-identification model based on body structure division. According to the method, the overall feature and the local feature are combined, the body structure information is fully utilized, the pedestrian featurewith higher discrimination is obtained through a bilinear fusion method, and the overall matching accuracy of the system is greatly improved.

Description

technical field [0001] The invention belongs to the technical field of computer vision pedestrian re-identification, in particular to a bilinear pedestrian re-identification network construction method based on body structure division. Background technique [0002] Pedestrian re-identification is a technology that uses computer vision technology to determine whether a specific pedestrian exists in an image or video sequence, and it is widely considered as a sub-problem of image retrieval. Given a monitored pedestrian image, retrieving the pedestrian image across devices and non-overlapping views plays an important role in criminal investigation, security monitoring, etc., and is an important part of the construction of a "safe city". [0003] There are two main categories of traditional pedestrian re-identification solutions: one is feature extraction, which is to extract robust feature representations based on pedestrians in surveillance videos; the other is distance measur...

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): G06K9/00G06K9/62G06N3/04
CPCG06V40/103G06V20/52G06N3/045G06F18/214Y02T10/40
Inventor 胡潇郭晓强周芸饶丰李下雨姜竹青门爱东
Owner ACADEMY OF BROADCASTING SCI STATE ADMINISTATION OF PRESS PUBLICATION RADIO FILM & TELEVISION
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