Tiếng Việt     English


We developed electronic hardware, firmware and software which automatically detect wildfire and determine fire risks based on environmental sensor data.

Camera-based wildfire detection: the software pre-processes and filters noises (frog, clouds) from color images, then converts to gray image to detect smoke contours. Depending on movement of smoke contours, we apply learning machine algorithms to decide whether wildfire occurs or not and sends SMS alert to users.

Sensor-based risk evaluator: the system collects environmental data such as rainfall, temperature, humidity, wind direction and wind speed via its sensors. The sensor data helps evaluate the fire risk and suggest preventive actions.

Picture. The firewatch station

  • Image-based smoke detection algorithms: SVM (Support Vector Machine)
  • Programming platform: OpenCV/C++
  • Environmental data logger:
    • MCU: STM32F7
    • Power: 24VDC/100mA
  • Connectivity:
    • Sensors: UART, analog inputs (4-20mA)
    • Cameras: RTSP, HTTP (GPRS/3G)
    • Software development: PHP web-application and MySQL database

Picture. The firewatch system