Interactive video surveillance by using environmental and mobile cameras
Itaru Kitahara · World Automation Congress · 2008
This paper introduces our researches aim to solve problems caused by increasing the number of surveillance cameras. ldquoViSErdquo is a semi-automatic surveillance system using multiple networked cameras to drastically reduce the burden of the operator who monitor the videos. The anonymizing video capturing system, ldquoStealth Visionrdquo can properly protect the privacy of target objects by fading out their appearance. We also research about a dynamically learning method for parametric eigenspace to realize an automated people identification system that accumulates knowledge of subjectspsila appearance as it captures surveillance videos.