Desarrollo de un sistema de visión computacional para la identificación de fallas superficiales en pavimento flexible en la ciudad de Trujillo

• Autores: Chávez Castro, Nataly Solange, Reyes Velásquez, Joseph Flamel
• Formato: tesis de grado
• Fecha de Publicación: 2019
• Institución: Universidad Nacional de Trujillo
• Repositorio: UNITRU-Tesis
• Lenguaje: español
• OAI Identifier: oai:dspace.unitru.edu.pe:20.500.14414/15980
• Enlace del recurso: https://hdl.handle.net/20.500.14414/15980
• Nivel de acceso: acceso abierto
• Materia: Visión artificial; Red neuronal convolucional; Procesamiento digital de imágenes; Visión computacional

The present research work aimed as main objective, to implement a computer vision_x000D_ system to identify surface faults in the flexible pavement using digital image_x000D_ processing and artificial intelligence fundamentals._x000D_ Four different models of convolutional networks were designed, with a data set_x000D_ of 7200 images, 1440 being for the training set, 360 for the validation set and 77 for_x000D_ the test. The model that presented the best indicators for the computer vision system_x000D_ was chosen, the results being as follows: 98.03% efficiency, 95.06% sensitivity and_x000D_ 95.63% specificity._x000D_ Finally, with the results obtained, the hypothesis proposed could be demonstrated,_x000D_ that a computer vision system allows the identification of surface faults in the flexible_x000D_ pavement