Forecasting Naturally Occurring Forest Fires using AI and Machine Learning in Wireless Sensor Networks

Wildfires are recurrent and sometimes devastating in regions of subtropical deciduous forests. Protecting forest areas are essential to ensure the health and vitality of the ecosystem. The goal of this work is to detect forest fires based on Artificial Intelligence (AI) and Machine Learning (ML) algorithms. The information obtained from the images combined with the forest spatial data will be used to make the detection. This article introduces an AI and ML based algorithm to forecast the forest fire in the wireless sensor network. The Convolutional Neural Network (CNN) algorithm is a better image classification algorithm for detecting fire. The experimental results illustrate that the proposed approach enhances the accuracy of fire detection. Furthermore, the proposed system using the CNN algorithm is compared to several ML algorithms such as Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), and K-Nearest Neighbor (KNN) algorithms. From the results, it is observed that the proposed algorithm minimizes the false positive and false negative ratio, and provides an accuracy of 97.35%.