Recognition of Humans and Their Activities for Video Surveillance

Abstract

Human Activity Recognition is an active area of research in computer vision with wide-scale applications in video surveillance, motion analysis, virtual reality interfaces, robot navigation and recognition, video indexing, browsing, HCI, choreography, sports video analysis, etc. The analysis of vision-based human activities in videos is an area with increasingly important consequences from security and surveillance to public place and personal archiving. Several challenges at various levels of processing-robustness against errors in low-level processing, view and rate-invariant representations at mid-level processing, and semantic representation of human activities at higher-level processing make this problem hard to solve. The task is challenging due to variations in motion performance, recording settings, and inter-personal differences. In this chapter, the authors explicitly address these challenges. They present a survey of existing work and describe some of the more well-known methods in these areas. They also describe their own research and outline future possibilities. Detailed overviews of current advances in the field are provided. Image representations and the subsequent classification processes are discussed separately to focus on the novelties of recent research. Moreover, the authors discuss the limitations of the state of the art and outline promising directions of research.