ESTG - Mestrado em Engenharia Informática - Computação Móvel
Permanent URI for this collection
Browse
Browsing ESTG - Mestrado em Engenharia Informática - Computação Móvel by advisor "Assunção, Pedro António Amado de"
Now showing 1 - 3 of 3
Results Per Page
Sort Options
- Object Detection in Omnidirectional ImagesPublication . Henriques, Francisco António Agostinho; Costa, Joana Madeira Martins; Assunção, Pedro António Amado de; Silva, Catarina Helena Branco Simões daNowadays, computer vision (CV) is widely used to solve real-world problems, which pose increasingly higher challenges. In this context, the use of omnidirectional video in a growing number of applications, along with the fast development of Deep Learning (DL) algorithms for object detection, drives the need for further research to improve existing methods originally developed for conventional 2D planar images. However, the geometric distortion that common sphere-to-plane projections produce, mostly visible in objects near the poles, in addition to the lack of omnidirectional open-source labeled image datasets has made an accurate spherical image-based object detection algorithm a hard goal to achieve. This work is a contribution to develop datasets and machine learning models particularly suited for omnidirectional images, represented in planar format through the well-known Equirectangular Projection (ERP). To this aim, DL methods are explored to improve the detection of visual objects in omnidirectional images, by considering the inherent distortions of ERP. An experimental study was, firstly, carried out to find out whether the error rate and type of detection errors were related to the characteristics of ERP images. Such study revealed that the error rate of object detection using existing DL models with ERP images, actually, depends on the object spherical location in the image. Then, based on such findings, a new object detection framework is proposed to obtain a uniform error rate across the whole spherical image regions. The results show that the pre and post-processing stages of the implemented framework effectively contribute to reducing the performance dependency on the image region, evaluated by the above-mentioned metric.
- Sportmove registro y monitorización de movimientos para la práctica deportivaPublication . Torres, Wilson Fernando Yandún; Marcelino, Luís Filipe Fernandes Silva; Assunção, Pedro António Amado de; Marcillo, DiegoEn el mundo existen más de 4.6 millones de aplicaciones móviles disponibles para sistemas operativos iOS y Android, orientadas a facilitar las actividades cotidianas de los seres humanos. El problema radica, en la escasa o nula existencia de aplicaciones móviles que integren un algoritmo que mida el rendimiento de los deportistas en base a la precisión de sus movimientos. La meta de esta investigación es explicar de manera detallada la propuesta, implementación, elaboración de pruebas y análisis de resultados de una aplicación móvil, enfocada en las necesidades de los usuarios deportivos (deportistas o entrenadores) que deseen evaluar, registrar y comparar el desempeño deportivo en base a la precisión de sus movimientos, empezando por la comparación de movimientos genéricos. Para desarrollar este trabajo se realizaron pruebas con distintos algoritmos, para hallar el que mejor se adapte a la investigación (Para evaluar los movimientos se utilizó R con la librería Similarity Measures). De manera paralela se captaron los movimientos de prueba, en un brazalete con un conjunto de sensores (Giroscopio, acelerómetro, sentido de orientación). Todos los datos de la investigación y el registro de los movimientos fueron almacenados en una base de datos. (MySql), para luego ser procesados por un software que realizó la integración de los datos, el sistema y el algoritmo mediante un framework de programación en el lenguaje PHP (Laravel). El trabajo demuestra las diferentes características de cada algoritmo y las dificultades que presentan para adaptarse a las necesidades de la investigación. Encontrar el algoritmo adecuado, permitió desarrollar una interfaz móvil, que asegura una mejor experiencia de usuario.
- Web-based Sports Scoring System for Trampoline Gymnastics CompetitionsPublication . Patrício, João Pedro Pereira; Assunção, Pedro António Amado de; Silva, Catarina Helena Branco Simões da; Marcelino, Luís Filipe Fernandes SilvaThis work arises within the context of the Master’s Project in Computer Engineering - Mobile Computing course, at Escola Superior de Tecnologia e Gest˜ao do Instituto Polit´ecnico de Leiria. The aim of this work is to describe the work carried out over several months, in partnership with the Trampolins Clube de Leiria. This is a local club, and as such, with somewhat limited resources, such as the competitions they manage and organize their judging solution were in need of improvements in terms of organization, speed, data acquisition, trust in the data, and resilience. This work begins by presenting the methodology used, followed by a contextualization of the sport, and a brief presentation of some software available in the market. Next, a comparison between the various current technological solutions is presented, as a means of gathering the necessary information to make informed decisions for the presentation of a solution. There is also a chapter with the details of the proposed solution, a description of the technologies used, and the implementation details. Finally, various execution scenarios, tests, and questionnaires are analyzed, concluding with the project’s findings. The main goal of the project was to propose an architecture for a solution to be made available in trampoline competitions, in order to minimize manual processes and perform calculations automatically, thus reducing human error and accelerating the entire end-toend process. There was also the objective of, during the development of the proposed solution, having it tested in actual events, including regional, national, and international competitions. The project was successfully completed, with the club’s main goals achieved. The experience gained in the interactions of requirement gathering was enriching for everyone. In the real-life context, the application’s performance and output were accepted in all possible categories.