Feature Extraction and Image Processing for Computer Vision, 4th Edition

9.2. Moving object detection

One of the main problems in detecting moving objects is that the insertion of a moving object in a scene does not obey the principle of superposition. Since the moving object obscures the background, there is no linear (filtering) approach which can separate the moving object from its background. As such we are left with a variety of approaches to achieve this task, as ever ranging from simple to complex and with different performance attributes.

9.2.1. Basic approaches

9.2.1.1. Detection by subtracting the background

The basic way to separate a moving object from its background is to subtract the background from the image, leaving just the moving object (the foreground). This is illustrated in Fig. 9.1 where ...