A Collaborative Pervasive Surveillance System COPS based on

A Collaborative Pervasive Surveillance System (COPS) based on low bit-rate video Supervised by Prof. Shueng Han Gary Chan Presented by Ho Chi Wang, Jody Tsang Yiu Wing, Wing Yiu Ching Chak, Billy

Overview of Presentation • • Tsang Yiu Wing, Wing Ho Chi Wang, Jody Yiu Ching Chak, Billy Background and Importance Work from Previous FYPs System Architecture System Functionalities Conclusion Quick COPS Demonstration Question and Answer

Background and Importance • Problems of Existing Video Surveillance System • • • Huge manipulation cost Intensive human involvement Location-restricted Collaborative and Pervasive • Our Group • • Tsang Yiu Wing, Wing Ho Chi Wang, Jody Yiu Ching Chak, Billy Designed system architecture Implemented core parts of COPS

Work from Previous FYPs • Not fulfill the requirements for the COPS – Only two components • • Centralized Server Simple Client Not effectively use the resources Not achieve a collaborative system – Not optimized settings • Not maintainable and flexible for extension – Code-and-Fix strategy – Highly coupled modules • Ho Chi Wang, Jody Tsang Yiu Wing, Wing Yiu Ching Chak, Billy Lack of project management and documentation

System Architecture – Overview • Our COPS - Distributed System Architecture was adopted – – – Real-time Video Streaming Server Application-Level Multicast Server Video Storage Server Web-based Image Server Dialogic Board Server – Intelligent Client Ho Chi Wang, Jody Tsang Yiu Wing, Wing Yiu Ching Chak, Billy

System Architecture – Pictorial View Ho Chi Wang, Jody Tsang Yiu Wing, Wing Yiu Ching Chak, Billy

System Architecture – Server and Client Structure • Layered Model – 3 layers was introduced Streaming Server GUI Ho Chi Wang, Jody Tsang Yiu Wing, Wing Yiu Ching Chak, Billy Capture Intrusion Intelligent Client GUI Encoder Decoder Network • A set of well-defined interfaces between modules • Implemented as Dynamic Link Library (DLL)

System Functionalities – Streaming Server • Real-time Streaming Server • • Multiple camera supports Video encoding – H. 263+ • 30 fps for QCIF • 15 fps for CIF • • Video streaming – UDP, RTP and TCP Intrusion detection and event-driven alert service Yiu Ching Chak, Billy Ho Chi Wang, Jody Tsang Yiu Wing, Wing

System Functionalities – Intelligent Client • Client-side Real-time Monitor • • • Video decoding – H. 263+ UDP, RTP and TCP Intrusion detection alert • Alarming sound • Storage Server Video Playback • • Video sequence searching File Playback • • Multiple video format support H. 263+, RGB, YUV Yiu Ching Chak, Billy Ho Chi Wang, Jody Tsang Yiu Wing, Wing

System Functionalities – Graphical User Interface • User Friendly Interface • • • Multiple camera previews Detailed event log Information display – fps, bps Icon menu Intrusion detection region settings Detailed system settings Server Client Yiu Ching Chak, Billy Ho Chi Wang, Jody Tsang Yiu Wing, Wing

Conclusion – Major Accomplishments • Successful System Re-engineering • • • Tsang Yiu Wing, Wing Ho Chi Wang, Jody Yiu Ching Chak, Billy A whole system architecture is designed The system is modulized according to functionalities Changes to each modules can be isolated Functions-rich and high performance system is developed Good Project Management and Documentations

Conclusion – Future Extension • Enhancing Video Encoding and Decoding Scheme • • Improving the service in the network module • • Tsang Yiu Wing, Wing Ho Chi Wang, Jody Yiu Ching Chak, Billy H. 264 – real-time video streaming with DVD quality Support quality of service (Qo. S) Enhance security service of the system

Quick COPS Demonstration Tsang Yiu Wing, Wing Ho Chi Wang, Jody Yiu Ching Chak, Billy

Question and Answer Tsang Yiu Wing, Wing Ho Chi Wang, Jody Yiu Ching Chak, Billy

FYP Demonstration • Date • • Time • • 11: 00 AM – 12: 00 PM Venue • • 8, May 2004 (Saturday) Pervasive Lab (Room 4203) COPS Official Website • http: //mwnet. cs. ust. hk/cops THANK YOU