An improved camshift-based particle filter algorithm for face tracking

Authors:
Jun Wang;Jin-ye Peng;Xiao-yi Feng;Lin-qing Li;Dan-jiao Li
Affiliations:
Department of Electronics and Information, Northwestern Polytechnical University, Xi'an, Shaanxi Province, P.R. China;Department of Electronics and Information, Northwestern Polytechnical University, Xi'an, Shaanxi Province, P.R. China;Department of Electronics and Information, Northwestern Polytechnical University, Xi'an, Shaanxi Province, P.R. China;Department of Electronics and Information, Northwestern Polytechnical University, Xi'an, Shaanxi Province, P.R. China;Department of Electronics and Information, Northwestern Polytechnical University, Xi'an, Shaanxi Province, P.R. China
Venue:
IScIDE'11 Proceedings of the Second Sino-foreign-interchange conference on Intelligent Science and Intelligent Data Engineering
Year:
2011

Citing 5
Cited 0

Detecting Faces in Images: A Survey

IEEE Transactions on Pattern Analysis and Machine Intelligence
Real Time Face and Object Tracking as a Component of a Perceptual User Interface

WACV '98 Proceedings of the 4th IEEE Workshop on Applications of Computer Vision (WACV'98)
CamShift guided particle filter for visual tracking

Pattern Recognition Letters
A Framework of Face Tracking with Classification Using CAMShift-C and LBP

ICIG '09 Proceedings of the 2009 Fifth International Conference on Image and Graphics
Multi-cue-based CamShift guided particle filter tracking

Expert Systems with Applications: An International Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

To improve the poor performance of face tracking due to complex illumination, cluttered background and poses variations, an improved Camshift-based particle filter algorithm is proposed in this paper. First, under the particle filter framework, we propose a novel feature extraction method called the block rotation-invariant uniform local binary pattern (BriuLBP), and combine with color features to represent the appearance model of face in tracking tasks. Compared to the traditional LBP, the BriuLBP method has the advantage of exploring space structure to enhance the accuracy of tracking, in term of preserving the robustness to illumination and appearance variations. Furthermore, in order to maintain the balance between sampling effectiveness and diversity, and to avoid Camshfit iterative to end at local optimal, a method of adjusting the number of particles in Camshift according to the target changing is proposed. Experimental results show that the proposed method performs better than the traditional particle filter algorithm based on Camshift in both robustness and accuracy, and achieve faster and more robust tracking for face tracking.