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

A surveillance video pedestrian re-identification method based on imagenet retrieval

A technology for pedestrian re-identification and monitoring video, applied in the field of video analysis, can solve the problem of long training process, achieve the effect of simple and easy implementation, improve the accuracy and environmental adaptability, and improve the actual use value

Active Publication Date: 2018-10-26
WUHAN UNIV
View PDF4 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But a practical problem is that the training of large-scale deep learning network requires a huge labeled training set, and the training process is extremely long, which cannot be tolerated by the criminal investigation business with limited time to solve the case

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
  • A surveillance video pedestrian re-identification method based on imagenet retrieval
  • A surveillance video pedestrian re-identification method based on imagenet retrieval

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0042]In order to facilitate those of ordinary skill in the art to understand and implement the present invention, the present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the implementation examples described here are only used to illustrate and explain the present invention, and are not intended to limit this invention.

[0043] A large number of existing pedestrian re-identification studies are based on a single standard dataset composed of pedestrians. However, in practical applications, pedestrians are not separated from surveillance videos, but mixed with background and other foreground objects. Manually labeled It is impractical to separate pedestrians in a large amount of surveillance video, therefore, a practical person re-identification method should be able to directly process surveillance video instead of a single pedestrian image. Object detection in video itself is ...

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 surveillance video pedestrian re-identification method based on ImageNet retrieval, which transforms the pedestrian re-identification problem into the retrieval problem of an active target image library, thereby utilizing the powerful classification ability of ImageNet hidden layer features. The steps include: preprocessing the surveillance video, removing a large number of irrelevant static background videos in the video; using the motion compensation frame difference method to separate the moving objects in the dynamic video frame, constructing a pedestrian image library and organizing an index table; Align the size and brightness of the target pedestrian image; use the ImageNet deep learning network to train the hidden features of the target pedestrian image and the image in the image database, and perform image retrieval based on cosine distance similarity; associate videos containing recognition results in chronological order Aggregated into video clips that reproduce pedestrian activity trajectories. The method of the invention can better adapt to changes in illumination, viewing angle, attitude and scale, and effectively improves the accuracy and robustness of pedestrian re-identification results in a cross-camera environment.

Description

technical field [0001] The invention belongs to the technical field of video analysis, and relates to a surveillance video pedestrian re-identification analysis method, in particular to a surveillance video pedestrian re-identification method based on ImageNet retrieval. technical background [0002] When solving crimes, the public security often needs to track suspected targets from a large number of surveillance videos with scattered geographical locations, large coverage areas, and long time spans. The existing manual video inspection method is easy to miss the best time to solve the case due to low efficiency. The criminal investigation business urgently needs automated analysis. and retrieval technology support. In this context, pedestrian re-identification technology came into being. Pedestrian re-identification refers to the technology of automatically matching the same pedestrian object under the non-overlapping multi-camera images in the illuminated area, so as to ...

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): G06K9/00G06T7/13G06T7/194
CPCG06V20/40G06V20/41
Inventor 王中元邵振峰胡瑞敏梁超
Owner WUHAN UNIV
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