Tracking human under occlusion based on kernel density estimation
WANG Xuan-he, LIU Ji-lin
Institute of Information and Communication Engineering, Zhejiang Provincial Key Laboratory of
Information Network Technology, Zhejiang University, Hangzhou 310027, China
The blob model, color model and motion model were built by fully using the space information, the motion information and the color information. The moving foreground targets based on mixture Gaussian model were detected. The people body in stored model was clustered to acquire the blob model based on Epanechnikov kernel density estimation algorithm. Color density function using non-parametric kernel density estimation algorithm and motion density function using Gauss distribution were obtained. The posterior probability images were attained by estimating the foreground of the current frame using the color density function and the motion density function. Then the occluded people in the posterior probability images were segmented. Finally, the results after segmenting occluded people were the final tracking targets. Experimental results show that the proposed algorithm efficiently solves the problem of tracking people under occlusion
[1] KANG J, COHEN I, MEDIONI G. Tracking people in crowded scenes across multiple cameras[C]∥Proceedings of Asian Conference on Computer Vision. [S.l.]:[s.n.], 2004:157-168.
[2] MITTAL A,DAVIS L. M2tracker: a multiview approach to segmenting and tracking people in a cluttered scene[J]. International Journal of Computer Vision, 2003,51(3):189-203.
[3] OTSUKA K,MUKAWA N. Multiview occlusion analysis for tracking densely populated objects based on 2d visual angles[C]∥Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition.[S.l.]:[s.n.],2004:1-90.
[4] BlACK J,EllIS T,ROSIN P. Multiview image surveillance and tracking[C]∥Proceedings of Workshop on Motion and Video Computing, 2002: 169-178.
[5] KHAN S,SHAH M. Tracking groups of people in presence of occlusion[C] ∥The 4th PacificRim Symposium on Image and Video Technology.[S.l.]:[s.n.],2000:438-446.
[6] SMITH K,GATICAPEREZ D,ODOBEZ J M. Using particles to track varying numbers of interacting people[C] ∥Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.San Diego:IEEE,2005:962-970.
[7] HAN M, XU W, TAO H,et al. An algorithm for multiple object trajectory tracking[C]∥ Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition. [S.l.]:IEEE,2004: 864-871.
[8] ISARD M,MAC J C. Bramble: a bayesian multipleblob tracker[C] ∥Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition. [S.l.]:IEEE,2001,2: 34-41.
[9] COMANICIU D,MEER P. Mean shift analysis and application[C]∥Proceedings of the 7th IEEE International Conference on Computer Vision.[S.l.]:IEEE, 1999:1197-1203.
[10] STAUFFER C,GRIMSON W E L. Adaptive Background Mixture Models for Realtime Tracking[C]∥Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Fort Collins. \
[S.l.\]:\
[s.n.\],1999:154-167.
[11] COMANICIU D, RAMESH V,MEER P.The variable bandwidth mean shift and datadriven scale selection[C]∥Proceedings of the 8th International Conference on Computer Vision. [S.l.]:[s.n.],2001:438-445.
[12] COMANICIU D, MEER P. Meanshift: a robust approach toward feature space analysis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(4):603-619.
[13] DAVID W S,SAIN S R. Multidimensional density estimation[J].Elsevier B V, 2004:24(5):229-260.
[14] DUONG T, HAZELTON M L. Crossvalidation bandwidth matrices for multivariate kernel density estimation[J]. Scandinavian Journal of Statistics, 2005,32:485-506.
[15] YANG C,DURAISWAMI R,GUMEROV N,et al. Improved fast gauss transform and efficient kernel density estimation[C] ∥Proceedings of IEEE International Conference on Computer Vision.[S.l.]:IEEE,2003: 464-471.
[16] COMANICIU D, RAMESH V, MEER P. Realtime tracking of nonrigid objects using mean shift[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.[S.l.]:IEEE,Hilton Head Island,2000:142-151.