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

Twin network tracking method based on self-adaptive template updating

An adaptive template and twin network technology, applied in the field of computer vision, can solve the problems of low tracking accuracy and robustness, and achieve the effects of low target tracking accuracy and robustness, real-time tracking speed, and strong feature representation ability.

Pending Publication Date: 2020-12-25
BEIJING UNIV OF TECH
View PDF0 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Then solve the problem of low tracking accuracy and robustness in the face of occlusion and rapid deformation in the process of target tracking

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
  • Twin network tracking method based on self-adaptive template updating
  • Twin network tracking method based on self-adaptive template updating
  • Twin network tracking method based on self-adaptive template updating

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] Due to the low tracking accuracy and poor robustness of the twin tracking algorithm in the prior art on various types of attributes, in order to improve the robustness of the twin tracking algorithm, a twin network tracking algorithm based on adaptive template update is invented in this paper. The implementation process of the present invention is described in detail below.

[0036] refer to figure 1 , a twin network tracking method based on adaptive template update, based on a target tracking model, the tracking model includes three modules, a twin tracking module, a trajectory prediction module and a template update module.

[0037] The tracking process mainly includes the following steps:

[0038] 1. Cut out a target template with a size of 127*127 from the first frame of the video;

[0039]2. Starting from the second frame, based on the bounding box of the object in the previous frame, cut out a search area with a size of 255*255 from the current original video fr...

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 twin network tracking method based on self-adaptive template updating, and is used for solving the problems of low tracking precision and robustness when the problems of shielding, rapid deformation and the like are solved in a target tracking process. The method comprises the following steps: firstly, inputting a video frame to be tracked and a target template into a twin tracking module, and outputting a predicted target tracking bounding box; inputting the historical tracking information into a trajectory prediction module, and judging whether to start a templateupdating module or not; if the template updating module is started, the template updating module updates the target template used by the current frame, and the updated target template replaces the target template before updating to serve as the input of the twin tracking module; and repeating the process to complete video tracking. According to the method, the target template features can adapt tothe appearance change of the target, real-time tracking is realized, and the problems of low target tracking precision and robustness when the target is shielded or quickly deformed and the like areeffectively solved.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a target tracking method based on a twin network of adaptive template updating. Background technique [0002] Single-target visual tracking is one of the important research directions in computer vision applications, such as image understanding, video surveillance, human-computer interaction, and automatic driving. Its purpose is to only give an arbitrary object in the initial frame, and then estimate the position of the object in the video sequence. One of the main challenges in object tracking is how to robustly represent appearance models due to frequently changing objects, different hues, presence of distractors, and environmental factors. In single object tracking, almost all state-of-the-art trackers improve the performance of tracking algorithms in two ways. One is to design efficient algorithms, called discriminative or generative algorithmic models. The former...

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): G06T7/20
CPCG06T7/20G06T2207/10016G06T2207/20132G06T2207/20221G06T2207/20084
Inventor 杨金福李亚萍李智勇李明爱
Owner BEIJING 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