摘要 |
Video is processed by maintaining a background model for the video, detecting a target in the video, detecting if the target is a stationary target, and classifying the stationary target as an insertion in the background model or a removal from the background model. |