Forest Fire Detection using Convolutional Neural Networks (CNN)

Abstract

One of the most significant and essential resources is the forest because it features a variety of plant life, including herbs, trees, and bushes, as well as several animal species. These renewable resources are crucial to humanity in some way. Forest fires, the most common hazard to forests, severely devastate the ecology, and local ecosystem. To preserve forests from fires, early detection and preventive measures are required. The two most common existing approaches for human surveillance to accomplish early detection are Direct human monitoring and remote video surveillance. This study proposes a forest fire image identification approach using convolutional neural networks to detect fires automatically. Employing this technique decreases false alarms and provides accurate fire detection results. The contour approach can be used to test its capability to monitor both interior and outdoor applications utilizing computer vision.