Unmanned Aerial Vehicle Assisted Forest Fire Detection Using Deep Convolutional Neural Network

Disasters may occur at any time and place without little to no presage in advance. With the development of surveillance forecasting systems, it is now possible forebode most life-threatening formidable disasters. However, forest fires are among ones that still hard anticipate beforehand, technologies detect plot their courses development. Unmanned Aerial Vehicle (UAV) image-based fire detection systems can be a viable solution this problem. these automatic use advanced deep learning image processing algorithms core tuned provide accurate outcomes. Therefore, article proposed method based on Convolutional Neural Network (CNN) architecture using new dataset. Notably, our also uses separable convolution layers (requiring less computational resources) for immediate typical layers. Thus, making suitable real-time applications. Consequently, after being trained dataset, experimental results show identify within images with 97.63% accuracy, 98.00% F1 Score, 80% Kappa. Hence, if deployed practical circumstances, identification used as an assistive tool outbreaks, allowing authorities respond quickly deploy preventive measures minimize damage.