Detection of objects from video images using background subtraction
Hiroaki Kai -, Jungpil Shin, Hsien‐Chou Liao · International Journal of Digital Content Technology and its Applications · 2012
Abstract The purpose of this study is to detect objects from video images. It seems that this research investigates visitor’s interest in the shop, and is useful for marketing purposes. We used a method called background subtraction for the realization of the system to detect objects from video images. We implemented the system by making use of a characteristic for the background subtraction to become the noises generally when we extracted human. We used frame differences to distinguish humans and objects. And we used the result, to extract and identify objects objectively. As a result of experiment, when a human was alone the system worked well. When the humans are multiple, a problem of shadow and occlusion are left. In future, this system will be applied for surveillance and crime prevention devices.