Early Detection of Forest Fire Based on Unmaned Aerial Vehicle Platform

Abstract

As a kind of natural resources, forest is vital to our survival. Once the fire happening in forest is not found in time, it can cause a huge loss. Accurate real-time monitoring of forest fires is not only an important part of forest fire prevention, but also an important means to effectively control the spread of forest fires and reduce economic losses. Forest areas are often geographically widely distributed and inefficient in manual inspections. Therefore, we need a highly mobile early forest fire warning system to ensure the safety of forest resources. UAVs serve as a highly mobile inspection tool to meet inspection requirements of forest fires. Besides, the early flame characteristics are not obvious, and the single source fire detection is easy to false alarm. In view of this, we propose a forest fire detection method that uses optical images to detect smoke and infrared images to detect fire.