Convolutional neural networks for identification of forest fires in satellite images: a short narrative review

Abstract

Deforestation, a global phenomenon resulting in massive loss of forest areas, and forest fires, which are increasing in frequency and intensity due to climate change and human activity, present major challenges in managing and reducing these catastrophic events. Forests are essential for biodiversity and, representing about one third of the earth's land surface, require effective protection and conservation strategies as a matter of urgency. The effectiveness demonstrated by the models in detecting forest fires with satellite images is highlighted, allowing a faster response to emergencies. However, some limitations are pointed out, such as satellite capabilities and the need for high quality data to ensure the reliability of CNN model performance. This paper reviews recent advances in this field, highlighting the effectiveness of CNN-based models in identifying fires accurately and in a timely manner.