Color Features for Tracking Non-Rigid Objects

Robust real-time tracking of non-rigid objects is a challenging task. Color distributions provide an efficient feature for this kind of tracking problems as they are robust to partial occlusion, are rotation and scale invariant and computationally efficient. This article presents the integration of color distributions into particle filtering, which has typically been used in combination with edge-based image features. Particle filters offer a probabilistic framework for dynamic state estimation and have proven to work well in cases of clutter and occlusion. To overcome the problem of appearance changes, an adaptive model update is introduced during temporally stable image observations. Furthermore, an initialization strategy is discussed since tracked objects may disappear and reappear.