The System of Fire Detection

• Main Authors: CHAN,YA-LI, 詹雅莉
• Other Authors: LIN,YUNG-CHIEN
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/48622000131579802333

碩士 === 吳鳳科技大學 === 光機電暨材料研究所 === 104 === Due to the high frequency of fire occurrence in modern environments, how to prevent fire effectively is an important issue. The fire loss can be minimized by taking a suitable fire prevention. Early the fire prevention was performed by people; and later the flame sensors were developed to detect the smoke and temperature. However the diffusion rates of the smoke and the temperature are slower, it leads to the shortcomings of the time delay and lower coverage on detection. Visual flame detection system can improve these shortcomings. The visual flame detection system can immediately identify the flame pixels, so that it can early find the fire point to avoid the fire expansion. This thesis is to develop a modular robot with automatic fire monitoring; the main controller is AT89S52 produced by Atmel; and a wireless remote computer is designed as an automatic fire detection robot control system to remotely control and immediately detect the temperature changes of environments. In the experiments, the robot is free to walk around, and detects the flame and the ambient temperature in an unknown environment. The human-machine interface is designed to control the robot action. The Sensirion's "SHT71" temperature sensor and the Logitech camera are used to detect temperature and to capture image around the environment. Finally, the flame recognition algorithm is employed to identify the suspicious fire zone. Keywords: robot, temperature sensors, fire detection, human-machine interface