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

A Cross-Camera Target Matching and Tracking Method Combined with Spatiotemporal Topology Estimation

A cross-camera and target matching technology, applied in the field of cross-camera target matching and tracking combined with spatio-temporal topology estimation, can solve the problems of not making good use of camera network topology information and difficulty in pedestrian feature matching

Active Publication Date: 2019-12-03
SUN YAT SEN UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] However, due to the change of viewing angle of each camera in the surveillance network, the influence of light, etc., it is very difficult to match pedestrian features under different cameras.
Most existing methods only consider the appearance representation of pedestrian objects, and do not make good use of the topological information between camera networks

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 Cross-Camera Target Matching and Tracking Method Combined with Spatiotemporal Topology Estimation
  • A Cross-Camera Target Matching and Tracking Method Combined with Spatiotemporal Topology Estimation
  • A Cross-Camera Target Matching and Tracking Method Combined with Spatiotemporal Topology Estimation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0075] The pure appearance feature matching does not consider the spatial topological relationship and time correlation constraints of the multi-camera network. If the space-time constraints are incorporated into the framework of target tracking, then the target matching between multiple cameras can obtain space-time clues, which can Greatly improve the accuracy of matching and tracking. Therefore, the present invention proposes a method combining target tracking and matching with topology estimation. Firstly, the current topological relationship is unsupervisedly established by using the current target matching result, and then the appearance model is combined with space-time constraints in the Bayesian framework. In turn, topological relationships are used to provide spatiotemporal clues for target matching, and continuous iteration makes the system stable, thereby achieving better multi-camera collaborative tracking.

[0076]Topology estimation mainly includes two aspects: ...

Embodiment 2

[0137] In order to verify the effectiveness of the algorithm, this embodiment collects the video data of a monitoring network consisting of five cameras in the building where the laboratory is located as the experimental object, and compares the experimental results with the real results of manual labeling, and then Two indicators, Precision and Recall, are used to characterize the optimization effect of the iterative algorithm. in,

[0138]

[0139]

[0140] Table 1 shows the changes in the precision rate and recall rate during the algorithm iteration process. It can be seen that the first iteration, that is, the initial state, is not effective because topology information is not used. As the algorithm continues to iterate, the monitoring network The topological information is effectively used, and the effect has been significantly improved. After that, the topological structure tends to be stable, so the effect also tends to be stable. Simultaneously, image 3 A comp...

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 mainly provides two points, according to one point, a target matching and tracing algorithm based on feature matching is researched, multiple kinds of apparent features with complementarity are adopted to set up a matching model, and the multiple feature matching results are fused at the decision level; according to the other point, an unsupervised topology estimation algorithm is provided so that a system can automatically set up the space-time topology relation of a monitoring network in the matching and tracing processes, and the matching and tracing accuracy is greatly improved by means of space-time topology constraint. The cross-camera target matching and tracing method has high robustness to interference caused by blocking, environment and illumination changes in the cross-camera target matching process and is beneficial for robust synergistic tracing of the target in the multi-camera video monitoring system.

Description

technical field [0001] The invention relates to the field of image tracking, and more specifically, relates to a cross-camera target matching and tracking method combined with spatio-temporal topology estimation. Background technique [0002] As people pay more and more attention to public security issues, as well as the country's planning and deployment of building safe cities and smart cities, video surveillance, as an effective security strategy, has been widely used in all aspects of society. Among them, target tracking has become a research hotspot in the field of computer vision as a key technology. Traditional video monitoring methods require manual processing and analysis of the collected video image sequences, while manual monitoring processing requires a lot of time and energy. At the same time, due to the expansion of monitoring scope and scale, manual monitoring and processing methods are difficult to achieve all-weather real-time operation. Therefore, people h...

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): G06T7/292G06T7/246G06T7/277H04N7/18
CPCG06T2207/10016G06T2207/20076G06T2207/30196G06T2207/30232H04N7/181
Inventor 郑慧诚林大钧许丹丹林梓健柯博
Owner SUN YAT SEN 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