Check patentability & draft patents in minutes with Patsnap Eureka AI!

Correlation filtering target tracking method and system based on sample differentiation learning

A technology of correlation filtering and target tracking, which is applied in the field of computer vision to achieve the effects of improving accuracy, reducing pollution, and increasing speed

Pending Publication Date: 2022-04-26
CHONGQING UNIV OF POSTS & TELECOMM
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

It is still a challenge to identify the reliability of different samples online and conduct differential model training

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
  • Correlation filtering target tracking method and system based on sample differentiation learning
  • Correlation filtering target tracking method and system based on sample differentiation learning
  • Correlation filtering target tracking method and system based on sample differentiation learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0079] The technical solutions in the embodiments of the present invention will be described clearly and in detail below with reference to the drawings in the embodiments of the present invention. The described embodiments are only some of the embodiments of the invention.

[0080] The technical scheme that the present invention solves the problems of the technologies described above is:

[0081] as attached figure 1 As shown, a correlation filter target tracking method and system based on sample differential learning includes the following steps:

[0082] 1. As attached figure 1(a) As shown in the benchmark tracker, the position and size of the target are obtained in the initial frame of the video, and a scale filter and positioning filter model are trained through preprocessing. The specific steps are:

[0083] 1) Obtain the state (including position and size) of the target in the initial frame from the video as a candidate area;

[0084] 2) Perform multi-scale cropping ...

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 provides a correlation filtering target tracking method and system based on sample differentiation learning, and belongs to the technical field of computer vision. The method mainly comprises the following steps: acquiring the position and size of a target in a video initial frame, and training a scale filter and a positioning filter; obtaining a response diagram of the candidate area of each frame of input image and the filter model; calculating values of a plurality of sample reliability indexes according to the response diagram; the label labeler labels reliability labels for the training samples on the basis of the index values; and finally, selecting a sample learning rate matched with the reliability label to update the filter model. Aiming at the problem that unreliable training samples exist in a complex tracking scene, a tracker is guided to differentially learn different samples by sensing the reliability of the samples. According to the method, pollution of unreliable samples to the model can be reduced, model drifting is relieved, and the tracking performance of a correlation filtering tracker in a complex scene is improved.

Description

technical field [0001] The invention belongs to the technical field of computer vision, in particular to a visual target tracking method and system. Background technique [0002] As an important topic and research hotspot in computer vision, video moving target tracking requires the tracker to accurately track and locate the behavior of the target in the video. With the continuous update of the algorithm, the target tracking theory has become more and more perfect and has attracted extensive attention from academia and industry. It is currently used in sports broadcasting, security monitoring and unmanned aerial vehicles, unmanned vehicles, robots and other fields. However, in actual complex tracking scenarios, moving objects in frames are usually accompanied by various influencing factors such as occlusion, illumination changes, scale changes, and motion blur, which lead to inaccurate tracking or even tracking failure. [0003] After more than 30 years of research and deve...

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): G06T7/246G06T5/20G06T5/40G06K9/62G06V10/774
CPCG06T7/251G06T5/20G06T5/40G06T2207/20081G06F18/214
Inventor 周丽芳李佳其李伟生王一涵冷佳旭
Owner CHONGQING UNIV OF POSTS & TELECOMM
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com