Analysis of video analytic architectures

Abstract

Video technology is changing rapidly due to the advancement in the sensor technology, transmission bandwidth, higher storage capability and availability of high resolution displays. Conventional video systems are being replaced by intelligent video systems to cater for automatic handling of the specific tasks and availability of high-volume data due to network of sensors. These intelligent video systems incorporate video analytics to meet the specific requirements of the user. Video analytics involves autonomous understanding of events by the machine in order to reduce the burden by removing the uninformative data for subsequent analysis and to augment the given task for monitoring, logging and action. There are numerous application areas of video analytics such as intrusion detection, crowd management, face recognition, unattended baggage detection, traffic management, video surveillance, object classification, detection & tracking, etc. The present work provides an overview of the available state-of-The-Art video analytics and the ongoing development in the area of video analytics. Framework for the development of video analytics algorithms and architectural issue has also been covered. © 2015 IEEE.