A partir de cette page vous pouvez :
Retourner au premier écran avec les étagères virtuelles... |
Résultat de la recherche
1 résultat(s) recherche sur le mot-clé 'Pavement cracks, distress, drone, Hahn moments, image processing dataset, Convolutional neural networks, deep learning.'
Affiner la recherche Faire une suggestion
PAVEMENT CRACK RECOGNITION SMART SYSTEM / Mohammed SALHI
Titre : PAVEMENT CRACK RECOGNITION SMART SYSTEM Type de document : projet fin études Auteurs : Mohammed SALHI, Auteur Année de publication : 2017 Langues : Anglais (eng) Mots-clés : Pavement cracks, distress, drone, Hahn moments, image processing
dataset, Convolutional neural networks, deep learning.Index. décimale : 1844/17 Résumé : The detection of pavement cracks in their early stages is considered to be one of
the most important challenges that stockholders face during the reparation of the
pavement condition. In other terms, the high severity cracks require huge budgets to
fix and inflict heavy losses to the vehicles. Hence the need to build an embedded
system that includes an autonomous unmanned aerial vehicle that sends data to the
cloud where it is processed to classify the pavement distress. This report explains the
different phases of the project. First, we select a new architecture of the deep learning
known by convolutional neural network based on Hahn moments, to be used for
image processing. Then, we reconstruct the INFOPAVE database to suit the study
purposes. After that, we proceed to the training and tests, to reach an accuracy of
100% in the training and 99.73% in the test. Later, we assemble an open hardware
drone that replies to the needs of the project in terms of current specifications and
perspectives in the future. Finally, the last part in the project focuses on the
communication between the UAV and the stakeholders. The system components can
be used separately in other such as studying the behavior of the pavement or using
the UAV for collecting another type of data.
PAVEMENT CRACK RECOGNITION SMART SYSTEM [projet fin études] / Mohammed SALHI, Auteur . - 2017.
Langues : Anglais (eng)
Mots-clés : Pavement cracks, distress, drone, Hahn moments, image processing
dataset, Convolutional neural networks, deep learning.Index. décimale : 1844/17 Résumé : The detection of pavement cracks in their early stages is considered to be one of
the most important challenges that stockholders face during the reparation of the
pavement condition. In other terms, the high severity cracks require huge budgets to
fix and inflict heavy losses to the vehicles. Hence the need to build an embedded
system that includes an autonomous unmanned aerial vehicle that sends data to the
cloud where it is processed to classify the pavement distress. This report explains the
different phases of the project. First, we select a new architecture of the deep learning
known by convolutional neural network based on Hahn moments, to be used for
image processing. Then, we reconstruct the INFOPAVE database to suit the study
purposes. After that, we proceed to the training and tests, to reach an accuracy of
100% in the training and 99.73% in the test. Later, we assemble an open hardware
drone that replies to the needs of the project in terms of current specifications and
perspectives in the future. Finally, the last part in the project focuses on the
communication between the UAV and the stakeholders. The system components can
be used separately in other such as studying the behavior of the pavement or using
the UAV for collecting another type of data.
Réservation
Réserver ce document
Exemplaires
Code barre Cote Support Localisation Section Disponibilité 1844/17 1844/17 MOH Texte imprimé unité des PFE PFE/2017 Disponible