Evaluación automática de daños pos-terremoto mediante UAVs

This work proposes a method of detecting collapsed buildings after an earthquake with low altitude aerial images, which are captured by an unmanned aerial vehicles (UAVs) and classified with a convolutional neural network (CNN). Different from the conventional methods that apply the satellite images or the high-altitude UAV for the coarse disaster evaluation over a large area, the purpose of this investigation is to achieve higher precision with the low altitud images. This thesis shows the advantages of CNN in comparison with other types of neural networks according with image recognition. The UAV images have unique advantages, such as stronger flexibility and higher resolution. The building damage was classified into five levels according with the European Macroseismic Scale 1988 (EMS-98) [1], instead of a binary classification of only collapsed and non-collapsed buildings as has been investigated by multiple authors. This damaged building mapping must be essential for the fast emergency response in Colombia cities.

• Autores: Rico Pérez, Julián Ricardo
• Fecha de publicación: 2019
• Institución: Universidad de los Andes
• Repositorio: Séneca: repositorio Uniandes
• Idioma: spa
• OAI Identifier: oai:repositorio.uniandes.edu.co:1992/43912
• Acceso en línea: http://hdl.handle.net/1992/43912
• Palabra clave: Edificios - Efectos de terremotos - Investigaciones; Riesgo sísmico - Investigaciones - Colombia; Análisis de imágenes - Investigaciones; Inteligencia artificial - Investigaciones; Redes neurales (Computadores) - Investigaciones; Drones - Investigaciones; Ingeniería
• Rights: openAccess
• License: https://repositorio.uniandes.edu.co/static/pdf/aceptacion_uso_es.pdf