Video Surveillance Analysis as a Context for Embedded Systems and Artificial Intelligence Education

Video surveillance analysis is an exciting, active research area and an important industry application.It is a multidisciplinary field that draws on signal processing, embedded systems, and artificial intelligence topics, and is well suited to motivate student engagement in all of these areas.This paper describes the benefits of the convergence of these topics, presents a versatile video surveillance analysis process that can be used as the basis for many investigations, and presents two template exercises in tracking detected targets and in evaluating runtime efficiency.The processing chain consists of detecting changes in a scene and locating and characterizing the resulting targets.The analysis is illustrated for targets in outdoor scenes using a variety of classification features.Also, sample code for processing is included.