Pedestrian re-identification method based on human skeleton mutual learning

A pedestrian re-identification and skeleton technology, applied in the field of pedestrian re-identification, can solve the problems of unreliable identity verification and unresolved problems of people's re-identification, and achieve accuracy improvement, accurate matching of local features, and high recognition accuracy Effect

Active Publication Date: 2019-07-09
SHANGHAI UNIV OF ENG SCI
View PDF9 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Despite the best efforts of computer vision researchers over the past decade, the problem of person re-identification remains largely unsolved
Especially in a busy environment monitored by remote cameras, relying on biometrics such as face and gait to authenticate people is unreliable

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
  • Pedestrian re-identification method based on human skeleton mutual learning
  • Pedestrian re-identification method based on human skeleton mutual learning
  • Pedestrian re-identification method based on human skeleton mutual learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is carried out on the premise of the technical solution of the present invention, and detailed implementation and specific operation process are given, but the protection scope of the present invention is not limited to the following embodiments.

[0038] This application is mainly applicable to the fields of cross-camera tracking and pedestrian search, such as figure 1 As shown, first bottom-up 2D pose estimation of pedestrians is performed, and then joint detection and pedestrian skeleton modeling are performed on the results.

[0039] Such as image 3 As shown, when performing joint detection, input the image into the CNN network framework, use the deep learning neural network and convolution operation to detect a single color map, output the heat map of each joint point of the human body, and use the peak value to represent...

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 pedestrian re-identification method based on human skeleton mutual learning, which comprises the following steps: (1) estimating pedestrian postures and skeletons by adopting a bottom-to-top method, marking articulation points of pedestrians in the process, carrying out local segmentation on the pedestrians by using an articulation point segmentation method, and executing local block matching; (2) estimating the joint point distance on the basis of a pedestrian 2D skeleton through a bottom-up method to learn global skeleton information, and executing global skeletonmatching; and (3) mutual learning is carried out by adopting biological feature-based local block matching and overall skeleton matching, classification loss and measurement loss are trained respectively, and the obtained mutual learning loss is shared in global feature matching to serve as priori experience to guide and correct global matching errors. Compared with the prior art, the method has the advantages of few influence factors, improved local matching accuracy, higher recognition accuracy and the like.

Description

technical field [0001] The invention relates to a pedestrian re-identification method, in particular to a pedestrian re-identification method based on human skeleton mutual learning. Background technique [0002] Person re-identification, a problem faced in person matching on non-overlapping cameras, has received increasing attention in recent years due to its importance in implementing automated surveillance systems. In many applications, such as cross-camera tracking and pedestrian search, it is desirable to recognize a person from a group of people based on appearance information. However, due to low resolution, motion blur, changes in view and individual appearance lighting, it is very challenging to construct a differentiated representation that adapts to different camera conditions, so in multi-camera systems, the matching of non-overlapping camera views has become more and more important. people's attention. For example, the behavior of a person in a large area of ​...

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
IPC IPC(8): G06K9/00G06N3/04G06N3/08
CPCG06N3/08G06V40/10G06N3/045
Inventor魏丹王子阳胡晓强罗一平
OwnerSHANGHAI UNIV OF ENG SCI