基于颜色与梯度方向联合直方图的Mean Shift跟踪

戴士杰; 齐金尧; 易丹; 邵猛; 李伟超

doi:10.3788/OPE.20152313.0459

您当前的位置：

首页 >

文章列表页 >

基于颜色与梯度方向联合直方图的Mean Shift跟踪

更新时间：2020-08-13

- 基于颜色与梯度方向联合直方图的Mean Shift跟踪
- Mean Shift tracking algorithm using joint histogram of colors and oriented gradients
- 光学精密工程 2015年23卷第10z期页码：459-465
- 作者机构：
  
  河北工业大学机器人及自动化研究所天津,300130
- 作者简介：
- 基金信息：
  
  国家自然基金委员会与中国民用航空局联合资助项目(No.V1433117)();河北省自然科学基金资助(No.F2012202041)();天津市民用航空器适航与维修重点
- DOI：10.3788/OPE.20152313.0459
  中图分类号： TP391.4
- 收稿日期：2015-03-15，
  
  修回日期：2015-04-10，
  
  纸质出版日期：2015-11-14
- 稿件说明：
移动端阅览
戴士杰, 齐金尧, 易丹等. 基于颜色与梯度方向联合直方图的Mean Shift跟踪[J]. 光学精密工程, 2015,23(10z): 459-465

DAI Shi-jie, QI Jin-yao, Yi Dan etc. Mean Shift tracking algorithm using joint histogram of colors and oriented gradients[J]. Editorial Office of Optics and Precision Engineering, 2015,23(10z): 459-465
戴士杰, 齐金尧, 易丹等. 基于颜色与梯度方向联合直方图的Mean Shift跟踪[J]. 光学精密工程, 2015,23(10z): 459-465 DOI： 10.3788/OPE.20152313.0459.

DAI Shi-jie, QI Jin-yao, Yi Dan etc. Mean Shift tracking algorithm using joint histogram of colors and oriented gradients[J]. Editorial Office of Optics and Precision Engineering, 2015,23(10z): 459-465 DOI： 10.3788/OPE.20152313.0459.

摘要

由于单一的基于方向纹理特征的模板匹配跟踪算法不能准确跟踪目标旋转

无法应对光变、目标旋转以及部分遮挡的复杂场景下的跟踪

本文在mean shift算法框架下提出一种基于颜色与梯度方向联合直方图的跟踪算法

该算法通过扩展直方图的维度将颜色与梯度方向特征融合到一个直方图中构成联合直方图。由光照变化模型计算得出光变因子表征光照变化程度

并作为两种特征的比例系数对联合直方图归一化。当光变因子小于某个阈值则选取颜色作为主特征

否则以梯度方向为主特征

以便充分发挥颜色对旋转鲁棒以及梯度方向对光照变化鲁棒的优势。通过将目标分为若干子区域单独跟踪

由匹配相似程度较大的子区域标记目标的位置解决部分遮挡问题。实验结果表明

提出的算法很好地处理了光照变化、目标旋转以及部分遮挡的复杂场景跟踪。

Abstract

The single orientation feature gets ineffective when tracking rotational object. For the sake of fulfilling object tracking in the scenes of illumination variation

rotation and partial occlusion

a joint histogram of colors and oriented gradients based Mean Shift tracking algorithm is presented. The joint histogram fuses the features of colors and oriented gradients by expanding the dimensions of the histogram. In this algorithm

IVF(illumination variation factor) is calculated by the model of illumination

which detects the degree of illumination variation. In the process of tracking

if IVF is below a certain threshold

color feature is taken as the principal feature so that the algorithm is robust to rotation

otherwise choose oriented gradients as principal feature for its robustness to illumination variation. The problem of partial occlusion can be figured out by dividing the object template into sub areas

and the position of object can be determined by the area that obtains largest similarity coefficient. The presented algorithm shows good performance in the experiments when dealing with complex scenes such as target rotation

illumination change and partial occlusion.

关键词

Keywords

references

LI S X, CHANG H X, ZHU C F. Adaptive pyramid mean shift for global real-time visual tracking[J]. Image and Vision Computing, 2010, 28(3):424-437.

WANG F, YU S, YANG J. Robust and efficient fragments-based tracking using mean shift[J]. AEU-international Journal of Electronics and Communications, 2010, 64(7):614-623.

COMANICIU D, RAMESH V, MEER P. Kernel-based object tracking[J]. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 2003, 25(5):564-577.

LEICHTER I, LINDENBAUM M, RIVLIN E. Mean shift tracking with multiple reference color histograms[J]. Computer Vision and Image Understanding, 2010, 114(3):400-408.

MAZINAN A H, AMIR-LATIFI A. Applying mean shift, motion information and Kalman filtering approaches to object tracking[J]. ISA transactions, 2012, 51(3):485-497.

WANG Y, CHEN F X, GUO H X. Kernel-based target tracking with spatial histogram and template drift correction[J]. Acta Automatica Sinica, 2012, 38(3):430-435.

孙中森, 孙俊喜, 宋建中, 等. 一种抗遮挡的运动目标跟踪算法[J]. 光学精密工程, 2007,15(2):267-271. SUN Z S, SUN J X, SONG J Z, et al.. Ant-occlusion arithmetic for moving object tracking[J]. Opt. Precision Eng., 2007, 15(2):267-271.(in Chinese)

YILMAZ A. Kernel-based object tracking using asymmetric kernels with adaptive scale and orientation selection[J]. Machine Vision and Applications, 2011, 22(2):255-268.

刘晴, 唐林波, 赵保军, 等. 改进的Mean shift目标跟踪算法[J]. 系统工程与电子技术, 2013, 35(6):1318-1323. LIU Q, TANG L B, ZHAO B J, et al.. Improved mean shift target tracking algorithm[J]. Systems Engineering and Electronics, 2013, 35(6):1318-1323.(in Chinese)

李远征, 卢朝阳, 高全学,等. 基于多特征融合的均值迁移粒子滤波跟踪算法[J]. 电子与信息学报, 2010, 32(2):411-415. LI Y ZH, LU ZH Y, GAO Q X, et al.. Particle filter and mean shift tracking method based on multi-feature fusion[J]. Journal of Electronics & Information Technology, 2010, 32(2):411-415.(in Chinese)

贾慧星, 张毓晋. 基于梯度方向直方图特征的多核跟踪. 自动化学报, 2009, 35(10):1283-1289. JIA H X, ZHANG Y J. Multiple kernels based object tracking using histograms of oriented gradients[J]. Acta Automatica Sinica, 2009, 35(10):1283-1289.(in Chinese)

浏览量

551

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于知识蒸馏的Transformer视觉跟踪器

针对RGBT跟踪的特殊属性的跨模态交互融合网络

基于显著性区域加权的相关滤波目标跟踪

基于双模型的相关滤波跟踪算法

基于双模型的相关滤波跟踪