| 5 | 0 | 1 |
| 下载次数 | 被引频次 | 阅读次数 |
针对视频背景复杂的情况，基于传统粒子滤波算法的跟踪方法不能实时准确跟踪目标的缺陷。本文结合传统粒子滤波算法和高斯-厄米特算法，并对视频进行预处理，为了提高跟踪速度，本文采用积分直方图描述人脸特征大大缩短了粒子的特征计算时间和权重更新时间。并且将颜色与边缘特征动态融合，并结合人脸形态的动态演变实时调整模板参数。实验验证了在应对类肤色、不同光照条件及其他复杂环境因素时，本文方法能够有效增强其稳定性和可靠性。
Abstract：Traditional particle filter-based tracking methods often fail to achieve precise tracking in complex video backgrounds.This article combines traditional particle filtering algorithm and Gaussian-Hermite algorithm to preprocess the video first.To improve tracking speed, this article uses an integral histogram to describe facial features, greatly reducing the feature calculation time and weight update time of particles. Additionally, dynamic fusion of color and edge features is implemented, combined with real-time adjustment of template parameters based on the dynamic evolution of facial morphology. The experiment verified that the method proposed in this paper can effectively enhance its stability and reliability when dealing with skin color, different lighting conditions, and other complex environmental factors.
[ 1 ] Welch G,Bishop G.An introduction to the Kalman filter[EB/OL].(2003-03-02)[2024-05-12].http://www.ceri.memphis.edu/people/smalley/ESCI7355/welch%20and%20bishop%20-%20kalman.pdf.
[ 2 ] Anderson B D,Moore J B.Optimal Filtering[M].New Jersey:Prentice-Hall,1979.
[ 3 ] Julier S J,Uhlmann J K.A new extension of the kalman filter to non lin ear systems[C]// Proc.Of Aero sense:The 11th Int.Symp.on Aero sp ace/Defense Sensing,Simulation and Controls,Orlando,FL,USA,1997.
[ 4 ] Isard M,Blake A.Condensation-conditional density propagation for visual tracking[J].International Journal of Computer Vision,1998,29(1):5-28.
[ 5 ] Yang D,Zhang Y L.An improved spatial histogramand particle filter face tracking[J].International Conference on Genetic and Evolutionary Computing,2015:257-267.
[ 6 ] Dou J F,Li J X,Zhang Z,et al.Face tracking with an adaptive adaboost-based particle filter[C]// IEEE Chinese Control and Decision Conference (CCDC),Xi′an,China,2012,24:3626-3631.
[ 7 ] 田天，陈刚.基于肤色和Gabor 纹理的粒子滤波人脸跟踪[J].计算机工程，2014,40(7):123-127.
[ 8 ] 南京大学数学系编.数值逼近方法[M].北京：科学出版社，1978.
[ 9 ] Jain A K,Hsu R L,Mohamed Abdel-Mottaleb.Face detection in color images[C]// IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(5):696-706.
[10] Swain M J,Ballard D H.Color indexing[J].International Journal Of Computer Vision,1991,7(1):11-32.
[11] Lin C.Face detection in complicated backgrounds and different illumination conditions by using YCbCr color space and neural network[J].Pattern Recognition Letters,2007,28(16):2190-2200.
[12] Valiska J,Marchevsky S,Kokoska R.Object tracking by color-based particle filter techniques in video sequences[C]// 24th International Conference,Radioelektronika,Bratislava,Slovak Republic,2014:1-4.
[13] Zhang M,Wang Q,He Z,et al.Bhattacharyya distance based kernel method for hyperspectral data multi-class classification[C]// Instrumentation and Measurement Technology Conference,Austin,TX,USA,2010:629-632.
[14] Zhu W,Levinson S.Edge orientation-based multi-view object recognition[C]// Proceedings 15th Internetionol Conference on Pattern Recognition,Barcelona,Spain,2000,1:936-939.
[15] 沈德海，侯建，鄂旭，张龙昌.基于Sobel的多方向算子模板边缘检测算法[J].现代电子技术，2015 (4) :91-93.
[16] 姚海涛.基于粒子滤波器的视频图像序列目标跟踪[D].武汉：武汉大学，2011.
基本信息:
DOI：
中图分类号:TP391.41;TN713
引用信息:
[1]周平平,张昊,李天平.基于高斯-厄米特粒子滤波算法的人脸跟踪方法[J].山东师范大学学报(自然科学版),2025,40(02):143-152.
基金信息: