Real-Time Kernel-Based Tracking in Joint Feature-Spatial Spaces

Changjiang Yang, Ramani Duraiswami, Ahmed Elgammal, Larry S. DavisDigital 2004

Abstract

An object tracking algorithm that uses a novel simple symmetric similarity function between spatially-smoothed kernel-density estimates of the model and target distribu-tions is proposed and tested. The similarity measure is based on the expectation of the density estimates over the model or target images. The density is estimated using radial-basis kernel functions which measure the affinity be-tween points and provide a better outlier rejection property. The mean-shift algorithm is used to track objects by itera-tively maximizing this similarity function. To alleviate the quadratic complexity of the density estimation, we employ Gaussian kernels and the fast Gauss transform to reduce the computations to linear order. This leads to a very ef-ficient and robust nonparametric tracking algorithm. The proposed algorithm is tested with several image sequences and shown to achieve robust and reliable real-time tracking. 1
BibTeX
@inproceedings{yang2004real,
  author = {Changjiang Yang and Ramani Duraiswami and Ahmed Elgammal and Larry S. Davis},
  title = {{Real-Time Kernel-Based Tracking in Joint Feature-Spatial Spaces}},
  booktitle = {Digital},
  year = {2004},
}

Loading related content...

Figure