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.