Publication: Solución IoT para promoción del turismo en Smart Cities mediante técnicas de Visión por Computador en entornos Cloud
Loading...
Official URL
Full text at PDC
Publication Date
2021
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citations
Exportar
Abstract
En el contexto del Internet de las Cosas (IoT, Internet of Things), y en particular de las Ciudades Inteligentes (Smart Cities), se ha diseñado y desplegado en distintos entornos Cloud una solución que permite a cualquier comercio/institución publicar, a través de distintos canales, información que pudiera resultar de interés a fin de promocionar los servicios que ofrece independientemente de la actividad que desarrolle.
Toda esta información podrá ser consultada por los viandantes que se encuentren frente a ellos mediante una fotografía donde se señale con el dedo índice al panel informativo que los identifica. Se ha desarrollado un modelo de detección de objetos capaz de identificar a partir de estas imágenes, en cooperación con servicios de reconocimiento de caracteres, la fuente de información solicitada a fin de recuperarla. El módulo de detección se ha diseñado utilizando tecnologías de Aprendizaje Profundo, concretamente el modelo YOLOv3, como método eficiente.
Con el fin de explotar las capacidades de este modelo, y de recuperar la información publicada de antemano, se ha provisionado en entornos sin servidor un servicio diseñado para atender a las distintas solicitudes de los usuarios de la solución que este trabajo propone.
La solución conceptual desarrollada permite verificar su validez en el contexto propuesto y bajo el paradigma IoT, quedando así disponible para su adaptación y despliegue en entornos reales.
In the context of the Internet of Things (IoT), and in particular of the Smart Cities, a solution has been designed and deployed in different Cloud environments that allows any company/institution to publish, through different channels, information that may be of interest to promote the services it offers, regardless of the activity it develops. All this information will be able to be consulted by the pedestrians in front of them by a photograph where they point with their index finger at the information panel that identifies them. An object detection model has been developed capable of identifying from these images, in collaboration with the character recognition services, the source of information requested to recover it. The detection module has been designed using Deep Learning technologies, specifically the YOLOv3 model, as an efficient method. In order to exploit the capabilities of this model, and to recover the nformation published in advance, a service designed to meet the requests of users of the solution proposed in this work has been provided through serverless environments. The conceptual solution developed allows to verify its validity in the proposed context and under the IoT paradigm, thus remaining available for its adaptation and deployment in real environments.
In the context of the Internet of Things (IoT), and in particular of the Smart Cities, a solution has been designed and deployed in different Cloud environments that allows any company/institution to publish, through different channels, information that may be of interest to promote the services it offers, regardless of the activity it develops. All this information will be able to be consulted by the pedestrians in front of them by a photograph where they point with their index finger at the information panel that identifies them. An object detection model has been developed capable of identifying from these images, in collaboration with the character recognition services, the source of information requested to recover it. The detection module has been designed using Deep Learning technologies, specifically the YOLOv3 model, as an efficient method. In order to exploit the capabilities of this model, and to recover the nformation published in advance, a service designed to meet the requests of users of the solution proposed in this work has been provided through serverless environments. The conceptual solution developed allows to verify its validity in the proposed context and under the IoT paradigm, thus remaining available for its adaptation and deployment in real environments.
Description
Máster en Internet de las Cosas, Facultad de Informática UCM, Departamento de Ingeniería de Software e Inteligencia Artificial, Curso 2020/2021.