Development of fire detection and warning system based on video processing

Abstract

Detect and warning fire is essential in our life. Many warning system has been developed in order to fulfill people’s command. However, the system use traditional ways, that is using normal sensor to detect the event. And it takes a long time to process and give out the alert. By using Embedded Computer, and applying Computer Science, we can improve the quality as well as save economy for customers. We replace traditional temperature sensors and microprocessor by embedded computer Raspberry Pi B+. Then we using network module F2103 to connect computer to server. Using algorithm we studied in Computer Science class, in here is CNN for image processing and hang out result. Also, we use other software such as Django to design website and Flutter for mobile apps. Our object are able to have the prototype system capture and detect fire event and notify customers with warning message immediately when event occur. The system is simple that everyone can used and enable to operate in any condition.