%0 Journal Article
%T Guest Editorial
%A WeiQi YAN
%A Sabu Emmanuel
%A Mohan S Kankanhalli
%J Journal of Multimedia
%D 2011
%I Academy Publisher
%R 10.4304/jmm.6.1.1-2
%X Surveillance can function both as a deterrent to help prevent crime, as well as an investigative tool to aid us in identifying an interloper. Multimedia surveillance systems enable us to record evidence, and carry out analysis on the generated/stored multimedia data. Analysis on these generated/stored surveillance data could help to reduce human¡¯s response time under certain critical situations and also save life and property. However, surveillance data such as video captured on a daily basis may be of huge volume. Without proper management, it is very hard to analyse this huge amount of surveillance data. This special issue focuses on multimedia surveillance from the experiential point of view. The semantics in multimodal multimedia data are particularly emphasized. The objective of this special issue is to gather high quality research efforts that address the challenges of the emerging area of multimedia surveillance with a view to providing the readers with an overview of the state of the art in this field.  After several rounds of review, four papers were selected for publication in this special issue. The first contribution introduces a framework for an event driven system for surveillance. The purpose of this system is to enable thorough exploration of surveillance events. The system uses client-server web architecture to provide the scalability for further development of the system infrastructure. The system is designed to be accessed by surveillance operators who can review and comment on events generated by the event detection processing modules. The contributions are in the development of automated real-time switching of camera views to aid camera operators, and also in the detection of events of interest within a surveillance environment. This system provides a novel approach to the technological surveillance paradigm. The second contribution is concerned with detecting objects of interest and obtaining their clear visual appearances. The proposed algorithm detects foreground objects from video sequences with fog and then enhances those objects¡¯ visibility. In addition it can be used to detect new objects entering the scene as well as sojourned objects.  The proposed algorithm is tested with surveillance video under different fog conditions. The third article presents a powerful technique to detect a flame in a surveillance video. . The research contributions are towards the development of a real-time flame detection system to aid with fire safety precautions in many different environments. The fourth paper is about multiple objects tracking w
%K Special Issue
%K Multimedia Systems for Surveillance
%U http://ojs.academypublisher.com/index.php/jmm/article/view/4466