Examinando por Autor "Forero Aranzalez, Duvan Nicolas"
Mostrando 1 - 1 de 1
Resultados por página
Opciones de ordenación
- ÍtemDetección de FOD por medio de inteligencia artificial con inspecciones por UAV(Fundación Universitaria Los Libertadores. Sede Bogotá., ) Forero Aranzalez, Duvan Nicolas; Albarracin Carranza, Andres Sebastian; Lozano Tafur, CristianThe FOD are a problem that occurs in general aviation, both in commercial and military aviation. It is estimated that foreign objects that can affect aircraft cost the aviation sector more than 4 billion dollars annually in both incidents and incidents, mainly taking into account maintenance costs. Seeing this great problem that is generated worldwide in the aviation sector have been implemented various alternatives for prevention, some inefficient, and other economically unviable for certain airports, so it seeks to implement the most modern technologies that are currently available, with artificial intelligence can generate detection of these FOD training a deep learning convolutional neural network for through an automated drone with its established flight plan can perform inspections much more efficient and faster than the current inspections that focus on the use of human personnel to detect these objects. Having this project a result of detection of objects with a CNN of 83% with a training of 2236 images with 7,838 labels. Validating in field tests, recommending what type of drones offered by the market could be used for this analysis and likewise indicating the altitude of 2 to 5 meters needed to have an adequate accuracy of at least 58%, being private with its flight plan based on the international airport El Dorado in Bogota.