Pedestrian re-identification method based on human skeleton key point detection and unequal partition

A pedestrian re-identification and key point technology, applied in the field of pedestrian re-identification, can solve the problems of poor model generalization ability, high false detection rate, and uncorresponding human body area division

Pending Publication Date: 2019-10-15
SHANDONG UNIV +2
View PDF2 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Due to the different proportions of the human body in the input picture, the uniform partition method of the PCB will result in mismatched human

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 key point detection and unequal partition
  • Pedestrian re-identification method based on human skeleton key point detection and unequal partition
  • Pedestrian re-identification method based on human skeleton key point detection and unequal partition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0064] A pedestrian re-identification method based on human skeleton key point detection and non-equal partition, such as figure 1 shown, including the following steps:

[0065] (1) Prepare the pedestrian data set, which refers to a large number of pedestrian images in different scenes and with different scales; divide the pedestrian data set into a training set, a verification set and a test set, and use the LabelImg image labeling tool to label the training set;

[0066] (2) Send the marked pictures in step (1) to the ResNet50 network to obtain tensor A representing the characteristics of the human body;

[0067] (3) According to the relative invariance of the human skeleton area, the tensor A is non-uniformly partitioned according to the key points of the human skeleton, and tensor1-1, tensor1-2, tensor1-3, tensor1-4, tensor1-5, tensor1-6 are obtained, respectively Characterize the head, chest, abdomen, thighs, calves, feet of the human body;

[0068] (4) Connect tensor1...

Embodiment 2

[0076] A method for re-identifying pedestrians based on human skeleton key point detection and unequal partitioning according to Embodiment 1, characterized in that:

[0077] In step (1), if Figure 5As shown, using the LabelImg image annotation tool to annotate the training set refers to: download and install the LabelImg image annotation tool, use the built-in drawing rectangle function in the LabelImg image annotation tool, and frame the location of pedestrians and each pedestrian in the pedestrian image key points of the human skeleton. In the pedestrian dataset, the training set accounts for 70%, the validation set accounts for 10%, and the test set accounts for 20%.

[0078] Step (2), such as Image 6 As shown, the steps are as follows: Send the picture marked with the LabelImg picture labeling tool to the input layer of the ResNet50 network in the form of pixels, and after the 50-layer convolution of the ResNet50 network, discard the downsampling of the last block in ...

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 key point detection and unequal partitioning, and the method comprises the following steps: (1), preparing a pedestrian data set, dividing the pedestrian data set into a training set, a verification set and a test set, and marking the training set; (2) sending to a ResNet50 network to obtain a sensor A representing human body characteristics; (3) according to the relative invariance of the human skeleton area, carrying out non-uniform partitioning on the sensor A according to the key points of the human skeleton, and respectively representing the head, chest, abdomen, thighs, shanks and feet of the human body; (4) respectively connecting FC layers, and classifying the FC layers by using a Softmax classifier; and (5) during verification and testing, combining the n feature vectors to calculate similarity, thereby realizing pedestrian re-identification. The pedestrian re-identification method based onhuman body key point detection and non-equal partition has stronger generalization ability and higher accuracy.

Description

technical field [0001] The invention belongs to the technical field of pedestrian re-identification, and in particular relates to a method for pedestrian re-identification based on human skeleton key point detection and uneven partitioning. Background technique [0002] Person Re-identification (Person Re-identification), also known as pedestrian re-identification, or ReID for short, is a technology that uses computer vision technology to judge whether a specific pedestrian exists in an image or video sequence, and describes multiple different images of a person distributed in different physical locations. The matching process for overlapping camera views. [0003] This technique can be widely used to re-identify, track or search for a person previously observed at a certain point in time in the camera network. It is designed to make up for the visual limitations of the current fixed camera, and can be combined with pedestrian detection and pedestrian tracking technology, a...

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): G06K9/00G06K9/62
CPCG06V40/10G06F18/214G06F18/2415
Inventor 范继辉周莉杜来民邓国超白玥寅张松朱顺意巩志远陈建学周雨晨
Owner SHANDONG UNIV
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