Mean Shift Algorithm and its Application in Tracking of Objects

Author

Wen, Zhi-qiang ; Cai, Zi-xing

Author_Institution

Coll. of Inf. Sci. & Eng., Central South Univ.

fYear

2006

fDate

13-16 Aug. 2006

Firstpage

4024

Lastpage

4028

Abstract

Mean shift algorithm is recently widely used in tracking clustering, etc, however convergence of mean shift algorithm has not been rigorously proved. In this paper mean shift algorithm with Gaussian profile is studied and applied to tracking of objects. The imprecise proofs about convergence of mean shift are firstly pointed out. Then a convergence theorem and its rigorous convergence proof are provided. Lastly tracking approach of objects based on mean shift is modified. The results of experiment show the modified approach has good performance of object tracking applied to occlusion. The contributions in this paper are expected to further study and application in mean shift algorithm

Keywords

Gaussian processes; hidden feature removal; object detection; object recognition; optical tracking; Gaussian profile; convergence theorem; mean shift algorithm; object tracking clustering; occlusion; Clustering algorithms; Computer aided instruction; Convergence; Cybernetics; Density functional theory; Educational institutions; Image segmentation; Information science; Iterative algorithms; Kernel; Machine learning; Machine learning algorithms; Pattern recognition; Bhattacharyya coefficient; Convergence; Kernel function; Mean shift algorithm; Tracking of object;

fLanguage

English

Publisher

ieee

Conference_Titel

Machine Learning and Cybernetics, 2006 International Conference on

Conference_Location

Dalian, China

Print_ISBN

1-4244-0061-9

Type

conf

DOI

10.1109/ICMLC.2006.258803

Filename

4028776