As the name suggests, it can subtract or exclude the background portion of an image. Its output is a binary segmented image that essentially gives information about non-stationary objects in the image. There is a problem with this concept of finding the non-stationary part, since the shadow of a moving object can move and sometimes be classified in the foreground.
Popular background subtraction algorithms:
BackgroundSubtractorMOG: This is a Gaussian blend based background segmentation algorithm.
BackgroundSubtractorMOG2: It uses the same concept but the main advantage it provides is stability even when there is a change in brightness and better identification of shadows in frames.
Geometric multigrid: a statistical method is used and the pixel bayesin segmentation algorithm.
Original video frame:
Background of the subtracted video frame:
Thus, we saw the application of a background subtraction algorithm that detects motion, life in video frames.