%0 Conference Proceedings
%@nexthigherunit 8JMKD3MGPCW/3F2PHGS
%@nexthigherunit 8JMKD3MGPCW/46KUES5
%2 sid.inpe.br/mtc-m12e/2026/04.29.19.26.59
%4 sid.inpe.br/mtc-m12e/2026/04.29.19.26
%3 Deep_Neural_Network_Utilization_in_Wildfire_Hotspot_Classification_For_Drone-Obtained_Imagery.pdf
%@mirrorrepository urlib.net/www/2025/02.16.04.07.51
%8 10-13 Nov. 2025
%@issn 979-833155003-5
%A Jesus, Leandro Diniz de,
%A Dantas, Antonio,
%A Shiguemori, Elcio Hideiti,
%A Vijaykumar, Nandamudi Lakalapalli,
%B Latin American GRSS and ISPRS Remote Sensing Conference (LAGIRS)
%@secondarytype PRE CI
%C Foz do Iguaçu
%D 2025
%@secondarykey INPE--PRE/
%I IEEE
%S Proceedings
%T Deep Neura Network Utiization in Widfire Hotspot Cassification For Drone-Obtained Imagery
%X Early detection and rapid response are critical in forest fire scenarios to minimize environmental and economic damage. This work explores the use of the YOLO (You Only Look Once) version 11 neural network for detecting three key classes: dispersed smoke, dense smoke, and fire. These classes were defined to support a tracking strategy aimed at identifying the fire source. In a typical forest fire, dispersed smoke is often visible from a distance, followed by dense smoke as one approximates its origin, and eventually, the flames themselves. A dataset composed of images from the Atlantic Forest was collected for training. The best model achieved an 85% mask accuracy with a recall of 76%. The network successfully segmented the target classes with few false positives, even on data not seen during training.
%@area COMP
%@electronicmailaddress leandro.jesus@inpe.br
%@electronicmailaddress antonio.josivado@unifesp.br
%@electronicmailaddress elciohs@gmail.com
%@electronicmailaddress vijay.n@inpe.br
%@documentstage not transferred
%@group CAP-COMP-DIPGR-INPE-MCTI-GOV-BR
%@group
%@group
%@group COPDT-CGIP-INPE-MCTI-GOV-BR
%@usergroup simone
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Universidade Federal de São Paulo (UNIFESP)
%@affiliation Instituto de Estudos Avançados (IEAv)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@versiontype publisher
%@holdercode {isadg {BR SPINPE} ibi 8JMKD3MGPCW/3DT298S}
%@doi 10.1109/LAGIRS68367.2025.11414805