An Intelligent Fire Detection and Warning System Under Home Video Surveillance for Garments Factories in Bangladesh

Devastating fire is a great concern in home monitoring system. Though there are many systems with state-of-the- art in fire detection but false signal is still a challenge to avoid. Fire detection and its information using image processing and information technology respectively can be a great hope to reduce devastation of fire mishaps. In multistoried buildings like a garments factory, fire can be originated or smelled from one unit of the building or floor but same type of alarm will put people into confusion, which can lead to a sudden panic. A lack of proper alarming approach in fire alarm system can be a reason of stampede death in a crowded multistoried building like garments factory in Bangladesh. Fake fire is another concern in conventional fire alarm system. In this project, an intelligent fire detection and warning system under home video surveillance is aimed in order to detect fire from video camera data and warn the concerned people accordingly. The system’s design will undergo a low computational time taking fire detection approach using RGB color space and it will warn the people dynamically than conventional alarming unit. The system will be able to cope up with the present conventional fire detection approach under security surveillance system. The proposed fire detection approach will be tested under different scenarios to confirm the accuracy of the system.

Required Qualifications: The potential candidate is expected to have engineering/physics or similar/relevant technological background to work in this project. He/she must have a high passion to contribute to a real-world problem with their skills and knowledge by realizing the current scenario of fire detection and alarming system in garments industries in developing countries like Bangladesh. Skill on image processing is also desirable.

Contact: Dr. M. Abdur Razzak, Project Supervisor & Professor, Department of EEE, IUB; E-mail: razzak@iub.edu.bd; Phone: +880-1753872800