Automatic Video Segmentation Algorithm by Background Model and Color Clustering

摘要

In order to detect the object in video efficiently, an automatic and real time video segmentation algorithmbased on background model and color clustering is proposed. This algorithm consists of four phases:backgroundrestoration, moving objects extract,moving objects region clustering and post processing. The threshold of thebackground restoration is not given in advanced. It can be gotten automatically.And a new object region