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

Correlation filtering video tracking algorithm based on residual network and short-term visual memory

A technique of visual memory and correlation filtering, applied in the field of computer vision

Pending Publication Date: 2021-01-08
TIANJIN CHENGJIAN UNIV
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the above invention does not involve the combination of residual network and visual memory mechanism

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 video tracking algorithm based on residual network and short-term visual memory
  • Correlation filtering video tracking algorithm based on residual network and short-term visual memory
  • Correlation filtering video tracking algorithm based on residual network and short-term visual memory

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] specific implementation plan

[0041] The present invention will be further described below in conjunction with the examples, the purpose is only to better understand the contents of the present invention, therefore, the examples given do not limit the protection scope of the present invention.

[0042] see Figure 1-Figure 4, the present invention provides a short-term memory visual correction model (SMRN) tracking algorithm based on the residual network to try to maintain robustness and accuracy, similar to the human cognitive memory system, SMRN extracts features through ResNet, and then uses kernel correlation The filter models the tracking of the target. Scale correction via short-term visual memory modules. At the same time, an adaptive learning method that adjusts and updates the appearance of the model through the cognitive memory mechanism is proposed, including the following steps:

[0043] 1. The overall structure of the algorithm:

[0044] The structure ...

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 relates to a correlation filtering video tracking algorithm based on a residual network and short-term visual memory, and aims to solve the problem that a target is easy to lose in a tracking process in a complex environment. The method comprises the following steps: firstly, extracting interested deep features of an image through ResNet different layers, and only selecting featuresextracted by a convolution layer with the best specific effect to train a related filter to obtain a target position with the maximum response value; secondly, performing scale sampling and memory sampling on the basis of determining the position, building a short-term memory scale pyramid, building a scale correlation filter accordingly, and therefore accurate estimation of the target scale is achieved. And finally, comparing with other algorithms in a data set OTB100, and experimental results show that the proposed algorithm obtains considerable accuracy and tracking success rate, and adaptsto complex environments such as illumination, scale change and occlusion under the condition of keeping certain real-time performance.

Description

technical field [0001] The invention relates to a correlation filtering video tracking algorithm based on residual network and short-term visual memory, which belongs to the technical field of computer vision. [0002] technical background [0003] Visual object tracking is one of the most important parts in the field of computer vision, and has broad application prospects in robot vision, artificial intelligence monitoring, AR, etc. In practical tracking tasks, the target region is generally specified in the first frame and tracked in subsequent frames. Although object tracking technology has achieved a lot of achievements in the past half century, it still has research significance due to factors such as light changes, deformation, sudden changes in motion and occlusion. [0004] It is generally believed that in the cognitive mental memory model, the human memory system contains three independent components: sensory memory, short-term memory and long-term memory. Sensory ...

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 Applications(China)
IPC IPC(8): G06T7/246G06N3/04G06K9/62
CPCG06T7/246G06N3/045G06F18/214
Inventor 任红格史涛梁晨赵坚杜静娟戈文琪吴启隆胡鸿长王东辉崔胤
Owner TIANJIN CHENGJIAN 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