Fire Detection System Using Infrared Thermometer and Image Processing with Level Identification and SMS Notification

Fire is the most common disaster that can happen daily. More people die from fire than any other kind of disaster. However, if there is an early warning from fire detection, and every person knows how to evacuate properly and calmly, a fire doesn't have to be fatal. Developing and implementing a Fire Detection System using image processing with fire level identification and SMS notification is to provide timely SMS notification and to detect fire to alert nearby personnel or authorities about the fire situation. To achieve this, image processing techniques are incorporated using a HuskyLens AI Machine Vision Sensor to detect fire from live video streams. The system is also capable of classifying the level of fire based on the intensity of the flame and sending the notification depending on the level identified. Arduino MEGA 2560 was used as the hardware platform along with the HuskyLens camera. The Arduino IDE 1.8.19 software was installed and used along with the HuskyLens library, Adafruit FONA library, Adafruit MLX90614 library, and Adafruit BusIO library. With the implemented system, fire detection and level identification with SMS notifications can be achieved with high accuracy and in real time.