Abstract

Accurately modeling ob ject colors, and features in general, plays a critical role in video segmentation and analysis. Commonly used color models, such as global Gaussian mixtures, localized Gaussian mix- tures, and pixel-wise adaptive ones, often fail to accurately represent the ob ject appearance in complicated scenes, thereby leading to segmenta- tion errors. We introduce a new color model, Dynamic Color Flow, which unlike previous approaches, incorporates motion estimation into color modeling in a probabilistic framework, and adaptively changes model parameters to match the local properties of the motion. The proposed model accurately and reliably describes changes in the scene’s appear- ance caused by motion across frames. We show how to apply this color model to both foreground and background layers in a balanced way for efficient ob ject segmentation in video. Experimental results show that when compared with previous approaches, our model provides more ac- curate foreground and background estimations, leading to more efficient video ob ject cutout systems.1