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

Cross-scene target automatic identification and tracking method and application thereof

An automatic recognition and scene technology, applied in neural learning methods, character and pattern recognition, instruments, etc., can solve problems such as tediousness, lack of analysis and key target search, and huge workload

Pending Publication Date: 2021-05-14
GUANGZHOU UNIVERSITY
View PDF0 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Moreover, the automatic monitoring commonly used in the market can only simply determine and classify abnormal behaviors. After an event occurs, it is usually followed by manual tracking of subsequent targets. There is a lack of follow-up analysis of the event and the search for key targets. The workload is huge and cumbersome.

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
  • Cross-scene target automatic identification and tracking method and application thereof
  • Cross-scene target automatic identification and tracking method and application thereof
  • Cross-scene target automatic identification and tracking method and application thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0091] Such as figure 1 As shown, this embodiment provides a cross-scene target automatic identification and tracking method, including the following steps:

[0092] S1: In the video, the ROI of the region of interest of the tracking target (that is, the actor) is automatically given by the system or manually, and the target is tracked in multiple monitoring scenarios using the Siammask neural network;

[0093] In this embodiment, the specific steps for establishing the target tracking model include:

[0094] S11: Obtain the target area (template) and the search area (search) in the next frame from the target position;

[0095] S12: If figure 2 As shown, Siamese is used as the backbone of the Siammask network to extract the template region features and search region features, and perform a deep cross-correlation between the two to obtain a response graph;

[0096] The template area is obtained as follows: the ROI can be input manually or by the system, and the setMouseCall...

Embodiment 2

[0232] Such as Figure 13 As shown, a cross-scene target automatic identification and tracking system includes a camera installed in each monitoring scene, a system server, and a system client; the video of each monitoring scene is uploaded to the system server for analysis, processing and real-time monitoring, and then Push the video monitoring results to the system client for real-time display and control, so as to realize the automatic identification and tracking of key targets.

[0233] In this embodiment, the camera is used to collect video images of each monitoring scene of the system, and transmit the collected video streams to the system server through the network (wired or wireless);

[0234] In this embodiment, the system server is used to integrate and implement the cross-scene target automatic identification and tracking method of the present invention, mainly including target tracking module, target re-identification module, cross-scene target tracking module, key...

Embodiment 3

[0245] This embodiment provides a storage medium, the storage medium may be a storage medium such as ROM, RAM, magnetic disk, optical disk, etc., and the storage medium stores one or more programs. Scene target automatic recognition and tracking method.

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 cross-scene target automatic identification and tracking method and the application thereof. The method comprises the following steps: a tracking target is tracked in a monitoring scene; when the tracking target is lost, the extraction model detects the tracked actor; the Euclidean distance between the actor sequence feature and the tracking target feature is calculated, and doer re-identification is carried out; the actor detection and actor re-identification is performed on multiple paths of scenes, and cross-scene actor identification and tracking is performed; the FACENET convolutional network face recognition technology is used for recognizing the actor, and identity recognition of the tracking target is carried out; the crowd density is estimated by adopting CSRnet, computing resources are allocated according to a crowd density threshold value, and early warning is performed on the monitoring area exceeding the set threshold value; key targets are searched and matched in all video monitoring scenes according to the retrieval conditions, and all targets meeting the conditions are selected. According to the invention, the generalization ability of the actor re-identification model in different scenes is improved.

Description

technical field [0001] The invention relates to the technical field of intelligent identification and tracking, in particular to a cross-scene target automatic identification and tracking method and application. Background technique [0002] With the development of computer technology, network technology and image processing technology, many video surveillance sites have been upgraded from traditional manual monitoring to automatic monitoring. According to research, the current automatic video monitoring methods often only detect abnormal events qualitatively, and realize simple judgment and classification of abnormal events, but lack of quantitative analysis of the process of abnormal events or after events. Especially the automatic search and identification of key people (and things) across cameras in abnormal events (and emergencies), and the automatic detection of safety warnings in crowded areas of public places. Moreover, the automatic monitoring commonly used in the ...

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): G06K9/00G06K9/32G06K9/46G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V20/52G06V10/25G06V10/44G06N3/045G06F18/214
Inventor 伍冯洁陈江涛吕明瑞王婷玉陈晓斌何唐梅朱静韩晓英
Owner GUANGZHOU UNIVERSITY
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