AI-BASED DISASTER DETECTION AND VICTIM LOCALIZATION USING DRONE IMAGERY AND DEEP LEARNING

This paper presents an AI-based framework for detecting disasters and locating victims from drone (UAV) imagery using deep learning. The framework has six layers: data acquisition, image preprocessing, deep-learning detection (YOLOv8/YOLOv10), fusion and geo-mapping, communication, and a dashboard for decision support. It uses CLAHE image enhancement, Non-Maximum Suppression, Kalman-filter tracking and a pixel-to-GPS transformation to turn raw aerial images into geo-referenced, prioritized incident reports for rescue teams. A review of fifteen recent studies identifies integration as the main research gap. In a controlled simulation, mean geolocation error was 1.95 m at 30 m altitude and 2.56 m at 120 m, and Kalman filtering reduced tracking error from 3.33 m to 2.16 m. The paper also discusses limitations and future validation.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-17
DOI
https://doi.org/10.5281/zenodo.24316414
Primary Topic
Advanced Neural Network Applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

AI-BASED DISASTER DETECTION AND VICTIM LOCALIZATION USING DRONE IMAGERY AND DEEP LEARNING

Gowtami Kattunga
3 links
Zenodo (CERN European Organization for Nuclear Research)
Advanced Neural Network Applications
article

AI-BASED DISASTER DETECTION AND VICTIM LOCALIZATION USING DRONE IMAGERY AND DEEP LEARNING

Gowtami Kattunga
article en

Abstract

This paper presents an AI-based framework for detecting disasters and locating victims from drone (UAV) imagery using deep learning. The framework has six layers: data acquisition, image preprocessing, deep-learning detection (YOLOv8/YOLOv10), fusion and geo-mapping, communication, and a dashboard for decision support. It uses CLAHE image enhancement, Non-Maximum Suppression, Kalman-filter tracking and a pixel-to-GPS transformation to turn raw aerial images into geo-referenced, prioritized incident reports for rescue teams. A review of fifteen recent studies identifies integration as the main research gap. In a controlled simulation, mean geolocation error was 1.95 m at 30 m altitude and 2.56 m at 120 m, and Kalman filtering reduced tracking error from 3.33 m to 2.16 m. The paper also discusses limitations and future validation.

Zenodo (CERN European Organization for Nuclear Research)
GMR Institute of Technology (IN)
Advanced Neural Network Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon