Drone imagery forest fire detection and classification using modified deep learning model

With the progression of information technologies, unmanned aerial vehicles (UAV) or drones are more significant in remote monitoring environment. One main application UAV technology relevant to nature is wild animals. Among several natural disasters, Wildfires one deadliest and cause damage millions hectares forest lands resources which threatens lives animals people. Drones present novel features convenience include rapid deployment, adjustable wider viewpoints, less human intervention, high maneuverability. effective enforcement deep learning many applications, it used domain fire recognition for enhancing accuracy detection through extraction semantic from images. This article concentrates on design drone imagery classification using modified (DIFFDC-MDL) model. The presented DIFFDC-MDL model aims imagery. To accomplish this, designs a MobileNet-v2 generate feature vectors. For classification, simple recurrent unit applied this study. In order further improve outcomes, shuffled frog leap algorithm used. simulation outcome analysis system was tested utilizing database comprising non-fire samples. extensive comparison study referred that improvements over other recent algorithms.