GEI and TripletLoss-DenseNet based gait recognition method

A gait recognition and gait technology, applied in the field of computer vision, pattern recognition, and deep learning, can solve the problems of complex data processing steps and low precision

Active Publication Date: 2018-11-30
BEIJING UNIV OF TECH
View PDF6 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problems of low accuracy and complicated data processing steps in the existing gait recognition technology when dealing with cross-view gait recognition

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
  • GEI and TripletLoss-DenseNet based gait recognition method
  • GEI and TripletLoss-DenseNet based gait recognition method
  • GEI and TripletLoss-DenseNet based gait recognition method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

[0041] Video image preprocessing process: the preprocessing process is as follows image 3 shown.

[0042] Step S1, get GEI (gait energy map):

[0043] Step S1.1, using the foreground detection method ViBe to extract the outline of pedestrians in the video image. First, the background is extracted in the first few frames of the video for background modeling, and then the ViBe (a moving object detection algorithm) method is used to directly extract the binarized pedestrian outline image in each frame of the video, and the extraction process uses The random update strategy updates the background sample points.

[0044] In step S1.2, since the image processed in step S1.1 has noise, for example, some backgrounds are mista...

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 discloses a GEI and TripletLoss-DenseNet based gait recognition method. The method uses a gait energy image GEI as the input of the network, connects all the layers of the network in a dense connection manner, and calculates a loss value of the training by using a triple loss function. The loss value is optimized to back-propagate and update model parameters, and the network model istrained until the network model converges. After training through the network, the GEI is finally mapped to feature vectors represented by a one-dimensional array in a specific space S, and the Euclidean distances between the feature vectors are used to represent the similarity of pedestrians. The similarity is used to match the gait to recognize the identity of the person. Testing is performed on a DatasetB of a CASIA gait database, and the model has strong feature mapping ability, which proves that the method can train a gait-based recognition model with excellent performance by using few training samples, and has the advantages of cross-view recognition and few model parameters.

Description

technical field [0001] The invention relates to the fields of deep learning, computer vision and pattern recognition, in particular to a gait recognition method based on a gait energy image (Gait Energy Image, GEI) and TripletLoss-DenseNet. Background technique [0002] Traditional biometric technologies include: face, iris, fingerprints, etc., but unlike these traditional technologies, gait recognition technology has unique advantages such as long-distance, uncontrolled, difficult to imitate, and difficult to hide, which makes it more Broad application background, including scientific research, transportation, criminal detection, etc. [0003] In the previous gait recognition methods, firstly, the pedestrian profile is extracted from the video sequence, and its gait energy map GEI is calculated, and then the similarity between different GEIs is compared, and finally classified and recognized by KNN. What these methods learn is an identity-specific classification ability. ...

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): G06K9/00G06N3/08
CPCG06N3/08G06V40/25
Inventor 杨新武侯海娥冯凯
Owner BEIJING UNIV OF TECH
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
PatSnap group products

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