Real-Time Integration of Segmentation Techniques for Reduction of False Positive Rates in Fire Plume Detection Systems during Forest Fires

Governmental offices are still highly concerned with controlling the escalation of forest fires due to their social, environmental and economic consequences. This paper presents new developments a previously implemented system for classification smoke columns object detection deep learning-based approach. The study focuses on identifying correcting several False Positive cases while only obtaining small reduction True Positives. Our approach was based using an instance segmentation algorithm obtain shape, color spectral features object. An ensemble Machine Learning (ML) algorithms then used further identify objects, removal around 95% Positives, 88.7% (from 93.0%) rate 29 newly acquired daily sequences. model also compared 32 sequences public HPWREN dataset 75 attaining 9.6 6.5 min, respectively, average time elapsed from fire ignition first detection.