Forest Firefighting Using Drone and Artificial Intelligence

This paper delves into exploring the implementation of Artificial Intelligence within the context of forest firefighting, specifically with regards to detecting fires and smoke. A variety of computer vision techniques are investigated with the intention of achieving real-time accuracy and efficacy. The VGG-16 model, a pre-trained convolutional neural network (CNN), is utilized for image classification, while Faster R-CNN and YOLOV8 are employed for detecting fires and smoke. Performance evaluation is conducted in both virtual 3D environments and real-world settings, with the ultimate goal of identifying the most precise and effective method for real-time detection. The objective of this research is to enhance forest firefighting strategies and contribute to the preservation of both ecosystems and human lives.