Stereo Video Surveillance Multi-agent System: New Solutions for Human Motion Analysis

摘要

This article presents a distributed agent-basedsystem that can process the visual information obtained bystereoscopic cameras. The system is embedded within aglobal project whose objective is to develop an intelligentenvironment for location and identification within dependentenvironments that merges with other types of technologies.In this kind of environments, vision algorithms are verycostly and require a lot of time to produce a response, whichis highly inconvenient since many applications can requireaction to be taken in real time. A multi-agent system (MAS)can automate the process of analyzing images obtained bycameras, and optimize the procedure. This study presents a MAS that can process stereoscopic images to detect andclassify people by combining a series of novel techniques.The article shows in detail the combination of techniquesused to perform the detection process. The process can besubdivided into human detection, human tracking, and humanbehavior understanding. With the addition of a casebasedreasoning (CBR) model, the system can also incorporatereasoning capabilities. The system was tested underdifferent conditions and environments.