Unlock instant, AI-driven research and patent intelligence for your innovation.

Object tracking algorithm based on combination between sparse expression and prior probability

A priori probability, sparse representation technology, applied in the field of computer vision, can solve problems such as target tracking that cannot be completely solved

Inactive Publication Date: 2017-03-22
ANHUI UNIVERSITY
View PDF3 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] To sum up, although the target tracking based on the idea of ​​sparseness shows a good tracking effect, it still cannot completely solve the target tracking under complex backgrounds such as noise, rotation, occlusion, motion blur, illumination and attitude changes.

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
  • Object tracking algorithm based on combination between sparse expression and prior probability
  • Object tracking algorithm based on combination between sparse expression and prior probability
  • Object tracking algorithm based on combination between sparse expression and prior probability

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0077] 1. Target tracking algorithm based on L1 regularization

[0078] The target tracking method based on L1 regularization was first developed by Mei et al. [7] Proposed, the following is a brief introduction based on particle filtering, and then gives the framework of the L1 tracking algorithm.

[0079] 1.1 Particle Filter Framework

[0080] Particle filter essentially implements Bayesian filtering through non-parametric Monte Carlo simulation, that is, using a set of random samples with weights to approximate the posterior probability density of the state of the system. Given the set of observations z up to time t-1 1：t-1 ={z 1 ,z 2 ,...,z t-1}, the best state of the target at time t can be obtained by the maximum approximate posterior probability z t *=argminp(x t i |z 1：t ). where x t i Indicates the system state of the i-th sampled particle at time t, the posterior probability p(x t i |z 1：t ) can be obtained recursively by Bayesian theory,

[0081] p(x...

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 an object tracking algorithm based on combination between sparse expression and prior probability by aiming at shielding, rotating, attitude changing, and motion blurring problems. Importance of an object template is measured by the prior probability, and the importance of the template is introduced in a regularization model, and is used as a main foundation of template updating, and therefore a new candidate object sparse coefficient solving method is acquired. By comparing with various popular algorithms, the algorithm can reach the same or higher tracking precision on a plurality of testing video sequences. According to experiment results, in the various videos having illumination changing, the attitude changing, and the motion burring problems, especially having shielding and out-of-plane rotating problems, the algorithm can track the objects stably and reliably, and therefore the improvement is more suitable for processing the videos having the shielded target problem, the motion blurring problem, the attitude changing problem, and the out-of-plane rotating problem.

Description

technical field [0001] The present invention relates to the field of computer vision, and more specifically, the present invention relates to a target tracking algorithm based on the combination of sparse representation and prior probability. Background technique [0002] Object tracking is one of the hot research directions in the current computer vision research field, and has a wide range of practical applications, such as automatic monitoring, intelligent navigation, human-computer interaction, and military defense. Although online object tracking has made great progress after decades of development, there are still many problems that have not been completely resolved. challenge [1-3] . [0003] Target tracking algorithms can generally be divided into tracking algorithms based on discriminative models and tracking algorithms based on generative models. Among them, the tracking algorithm based on the discriminant model transforms the tracking problem into a classificat...

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): G06K9/32
CPCG06V10/245
Inventor 周健田猛
Owner ANHUI UNIVERSITY