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

Visual target tracking algorithm based on multi-dimensional confidence evaluation learning

A target tracking and confidence level technology, applied in the field of visual target tracking, can solve the problems of low target tracking accuracy and tracking loss, and achieve the effects of avoiding filter drift, good evaluation, and improving accuracy

Inactive Publication Date: 2021-11-19
XIAN UNIV OF TECH
View PDF0 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a visual target tracking algorithm based on multi-dimensional confidence evaluation and learning to solve the problems of low target tracking accuracy and tracking loss and re-detection in complex environments

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
  • Visual target tracking algorithm based on multi-dimensional confidence evaluation learning
  • Visual target tracking algorithm based on multi-dimensional confidence evaluation learning
  • Visual target tracking algorithm based on multi-dimensional confidence evaluation learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0061] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] The present invention is a visual target tracking algorithm process based on multi-dimensional reliability evaluation. figure 1 As shown, the specific method is:

[0063] Step 1, before the tracking begins, set the multi-dimensional assessment of the corresponding threshold and the fusion coefficient, and the multi-dimensional evaluation threshold includes the granary evaluation threshold η. 0 , Similarity evaluation threshold η 1 . Confidence evaluation fusion coefficient contains coefficient, similarity, and smoothness fusion coefficient (ξ o Ξ p Ξ s ); Set the template failure threshold; continuous sampling frame number m; filter template and learning rate initialization; set the GMM sample space size n and set the number of per-frame weights a. In order to improve confidence evaluation accuracy, this article is sampled for continuous M f...

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 visual target tracking algorithm based on multi-dimensional confidence evaluation learning, and solves the problems of low target tracking precision and tracking loss re-detection in a complex environment. According to the method, a mapping relation with adaptive learning parameters of a filter is established according to multi-dimensional confidence evaluation, and a high-confidence sample set is constructed and managed by adopting a Gaussian mixture model (GMM).According to the method, the learning step length of the filter is adaptively adjusted according to a confidence evaluation score in the tracking process, and the negative influence of samples with unreliable tracking results on filter parameter training is suppressed, so that filter drift is avoided, tracking performance is prevented from being reduced, and the accuracy and robustness of visual target tracking in a complex environment are improved.

Description

Technical field [0001] The present invention belongs to the field of target tracking vision, particularly relates to the visual target tracking algorithm based on confidence Multidimensional assessment study. Background technique [0002] With the growing demand for intelligent video surveillance applications, human-computer interaction and precise navigation and other related fields, people's real-time, accurate, robust object tracking put forward higher requirements. Related filtered target tracking algorithm is based on an efficient, reliable visual object tracking method. However, many interfering factors (such as deformation, occlusion, fast motion, etc.) will reduce the actual environment resulting filter performance training target tracking failure. [0003] Relevant existing filter tracking algorithm confidence assessment strategies over a single, simple apply only to a specific scene. Based on the confidence map correlation filter core (Kernel Correlation Filter, KCF) al...

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): G06T7/246G06K9/46G06K9/00
CPCG06T7/246G06T2207/30241
Inventor 史思琦李南廷马彦军郑莉平
Owner XIAN UNIV OF TECH
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