A Machine Learning-Based Early Forest Fire Detection System Utilizing Vision and Sensors' Fusion Technologies

The paper aims at utilizing machine learning (ML) towards designing an early warning forest fire detection system.With the aid of the Internet of Things (IoT) and smart edge computing, an embedded system that utilizes sensors' fusion technology, machine vision and ML to early detect forest fire has been proposed.Different from most of the proposed fire detection systems in the literature, which either utilize vision or sensors'-based approaches to detect the fire, the proposed system utilizes both approaches jointly, which in turn will make it more accurate for fire detection.Furthermore, this paper focuses on implementing the proposed system utilizing a smart edge node and discusses the incurred technical challenges and how they have been solved.