Target tracking algorithm based on state perception template updating

A template update, state-aware technology, applied in the field of target tracking in computer vision, can solve the problems of error accumulation, target offset, loss, etc., to achieve the effect of reducing background error, good convergence effect, fast and reliable template update

Pending Publication Date: 2021-06-29
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The current common template update methods are mostly simple combination of target features. This method can adapt to the deformation of the target to a certain extent, but at the same time it will also introduce the background information of the target in the current frame. This part of information is often unnecessary. In the past, it will cause errors to accumulate, which will also lead to target deviation and loss

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
  • Target tracking algorithm based on state perception template updating
  • Target tracking algorithm based on state perception template updating
  • Target tracking algorithm based on state perception template updating

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0022] The present invention will be further described below in conjunction with the accompanying drawings.

[0023] The invention proposes an improved Siamese template tracking algorithm based on adaptive template update, aiming at optimizing the adaptability of the target tracking algorithm to the problem of target deformation. The overall framework of the algorithm is as follows figure 1 shown.

[0024] The algorithm consists of three modules: basic tracking module, template updating module and state perception module. The basic tracking module mainly performs regression and classification on reliable objects in the search area; the state perception module aims to dynamically judge the current state of the target, and decide whether to enter the template update module; the template update module focuses on the template that can quickly and effectively carry out the target renew.

[0025] (1) Basic tracking module

[0026] Basic tracking modules such as figure 2 As sho...

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 an improved target tracking technology based on state perception template updating. When a target is tracked, the shape and size of the target in an image sequence can be changed along with the change of an angle between an image acquisition device and the target or the morphological change caused by the characteristics of the target, namely, the problem of deformation of the target is caused, and a certain challenge is brought to a target tracking algorithm. According to a target deformation problem, a Siamese target tracking technology based on state sensing template updating is provided based on a SiamRPN algorithm, and template updating in target tracking is improved from two aspects. On one hand, how to update the template is combined with the thought of meta learning, a template update network is constructed, and rapid and reliable template update is realized. On the other hand, when template updating is carried out, the target state is judged by constructing a long and short memory network, and the opportunity for updating the target template is determined. According to the method, the problem of target template updating in a long-time target tracking scene can be effectively solved.

Description

technical field [0001] The invention relates to the field of target tracking in computer vision. Aiming at the problem of target deformation in a long-term target tracking scene, an improved target tracking algorithm for state-aware template update is proposed. Background technique [0002] Visual object tracking technology is one of the important basic technologies of computer vision, and it has always been an important topic and research hotspot in the field of computer vision. It aims to estimate the position and size of the target in the subsequent video sequence by giving the initial position of the target in the initial frame of the video. At present, target tracking has received extensive attention in the field of computer vision research, and has a wide range of applications in the field of machine intelligence, including military guidance, automatic driving, human-computer interaction, intelligent monitoring, and intelligent transportation. [0003] During target t...

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 & AuthorityApplications(China)
IPC IPC(8): G06T7/246G06K9/46G06K9/62G06N3/04G06N3/08
CPCG06T7/246G06N3/08G06T2207/30241G06T2207/10016G06T2207/20081G06V10/44G06N3/048G06N3/044G06N3/045G06F18/241
Inventor贾海涛邓文浩王子彦贾宇明任利许文波田浩琨张诗涵
OwnerUNIV OF ELECTRONICS SCI & TECH OF CHINA