A UAV-based Forest Fire Detection Algorithm Using Convolutional Neural Network

Abstract

With the new development in Unmanned Aerial Vehicle (UAV) in recent years, UAV equipped with color and infrared cameras becomes a new tool for carrying out the forest fire detection and fighting missions due to its advantages of low price, high maneuverability, and easy use. In order to detect a potential fire in its early stage, a UAV-based forest fire detection method using a convolutional neural network method is proposed in this paper. The effectiveness of the proposed fire detection algorithm is verified by using simulated flames in an indoor experimental testbed.