A real-time two-dimensional target tracking method based on two-way feedback particle filter algorithm

A particle filter algorithm and two-way feedback technology, applied in computing, image data processing, instruments, etc., can solve the problems of particle degradation, dynamic target tracking, and large amount of particle filter calculations, so as to avoid particle shortage, narrow the gap, reduce The effect of calculation volume

Active Publication Date: 2018-06-01
NANJING WEIJING SHIKONG INFORMATION TECH CO LTD
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The technical problem to be solved by the present invention is that the existing particle filter has the defects of large amount of computation, particle degradation and inability to track dynamic targets

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
  • A real-time two-dimensional target tracking method based on two-way feedback particle filter algorithm
  • A real-time two-dimensional target tracking method based on two-way feedback particle filter algorithm
  • A real-time two-dimensional target tracking method based on two-way feedback particle filter algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0041] Such as figure 1 As shown, the real-time two-dimensional target tracking method based on the two-way feedback particle filter algorithm provided by the present invention includes the following steps:

[0042] Step 1, extract and record the four types of features of the two-dimensional target to be tracked in the tracking gallery, including color, texture, shape, and multi-scale feature points based on FAST, and then generate N particle processes corresponding to the two-dimensional target to be tracked;

[0043]Step 2, import the video frame of the video signal according to the time series, at t=0 moment, N particle processes independently use the color and texture features to carry out feature matching to the video frame, and calculate the particle weight of each particle process; at t>0 moment, If the effective weight at time t-1 < weight threshold Vpt, then N particle processes use the features used at time t-1 to perform feature matching on the video frame at time t...

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 bidirectional-feedback-particle-filter-algorithm-based real-time two-dimensional target tracking method. The method comprises: feature extraction is carried out and N particle processes are generated, a video frame is inputted, and feature matching is carried out based on a weight threshold Vpt selection feature; the number of effective particle processes is calculated and whether the number of the effective particle processes is larger than N / 5 is determined based on comparison; if the number of the effective particle processes is smaller than the N / 5, weight optimization is carried out to reduce distances among the N particle weights; and if the number of the effective particle processes is larger than N / 5, resampling calculation is carried out and the particle process with the high weight value is used for covering and replacing the particle process with the low weight value; the particle process with the maximum particle weight is selected as a tracking result of a two-dimensional target in the video frame at a t time; and then updating of four features, system noises and measurement noises is carried out. According to the invention, with the method, defects that the operation load is high, particle degeneracy is serious and a dynamic target can not be tracked according to the existing particle filter method can be overcome.

Description

technical field [0001] The invention relates to a two-dimensional target tracking method, in particular to a real-time two-dimensional target tracking method based on a two-way feedback particle filter algorithm. Background technique [0002] Particle filter is suitable for dealing with nonlinear and non-Gaussian state estimation problems, suitable for tracking in complex environments, and can recover tracking from transient failures. However, the most fatal disadvantage of conventional particle filtering is the large amount of computation. In order to obtain accurate state estimation, a large number of particles are required to describe the posterior probability distribution of the tracking scene, and as the estimated state dimension increases, the number of particles will increase exponentially. . Another disadvantage of conventional PF is the problem of particle degradation, that is, as the iterations continue, the variance of particle weights will continue to increase, ...

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 Patents(China)
IPC IPC(8): G06T7/246
CPCG06T2207/20024
Inventor 冯向文潘铭星赵金辉孙健杨佩星付俊国
Owner NANJING WEIJING SHIKONG INFORMATION TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products