Summary: BU CS TR99-015.v2, Dec. 1999 (revised in March 2000).
To appear in Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR 2000).
Estimation and Prediction of Evolving Color Distributions for Skin
Segmentation Under Varying Illumination
Leonid Sigal, Stan Sclaroff, and Vassilis Athitsos
Image and Video Computing Group - Computer Science Dept.
Boston University - Boston, MA 02215
A novel approach for real-time skin segmentation in video
sequences is described. The approach enables reliable skin
segmentation despite wide variation in illumination during
tracking. An explicit second order Markov model is used
to predict evolution of the skin color (HSV) histogram over
time. Histograms are dynamically updated based on feed-
back from the current segmentation and based on predic-
tions of the Markov model. The evolution of the skin color
distribution at each frame is parameterized by translation,
scaling and rotation in color space. Consequent changes
in geometric parameterization of the distribution are prop-
agated by warping and re-sampling the histogram. The