| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 9.06 MB | Adobe PDF |
Advisor(s)
Abstract(s)
No quotidiano das sociedades atuais, a sinistralidade automóvel permanece como uma das principais causas de morte a nível mundial, vitimando todos os anos milhões de pessoas e provocando danos físicos irreversíveis, que resultam em incapacidades totais ou parciais. Com o crescimento sustentado das tecnologias inteligentes de transporte, denominadas “ITSIntelligent Transportation Systems”, verifica-se que as mesmas são um excelente meio de atuaçãona prevenção de acidentes rodoviários, particularmente aquelas que auxiliam directamente o condutor dentro do veículo “ADAS – Advanced Driver Assistance System”. A introdução dos sistemas ADAS no mercado tarda em proliferar, devido a restrições como o custo e disponibilidade, já que normalmente a sua disponibilização é influenciada pelo foco comercial dos fabricantes de automóveis e está inerente à utilização de carros de gama alta ou à aquisição de extras dispendiosos. Considera-se também, que os requisitos tecnológicos associados a estas soluções são um entrave à sua implementação, uma vez que a idade média dos automóveis é elevada. O défice de alcance deste tipo de soluções pela população surge como motivação desta dissertação, para o desenvolvimento de um sistema capaz de ser acessível por todos. Esta dissertação tem como objetivo principal a demonstração de uma investigação na área dos sistemas inteligentes de transporte, sendo base para o desenvolvimento de um protótipo de um sistema evolutivo de assistência ao condutor, denominado Virtual Driver. A sinalização de trânsito é o principal meio de comunicação entre o condutor e estrada, definindo quais são as regras e perigos inerentes na sua utilização. No entanto, a sua não visualização por fatores como distração, falta de foco na condução ou falta de visibilidade, podem comprometer a segurança do próprio condutor e de outros presentes na via, resultando em acidentes rodoviários. Virtual Driver é uma solução de baixo custo que é executável através de dispositivos presentes no quotidiano, como o smartphone ou outros micro-computadores. Desta forma, permite, numa primeira fase, auxiliar o condutor através do reconhecimento de sinalização vertical de trânsito em tempo real, alertando-o através de informações áudio-visuais. O protótipo desenvolvido atua com base na sinalização vertical de trânsito Portuguesa, pretendendo-se, mais tarde, abranger outras realidades. Numa fase final, será apresentada a avaliação do desempenho da solução, relativamente aos processos de deteção e reconhecimento de sinalização, evidenciando a utilização da metodologia escolhida.
In today's society, road accidents remain one of the leading causes of death worldwide, killing millions of people each year and causing irreversible physical damage, resulting in total or partial disability. With the sustained growth of Intelligent Transportation Systems (ITS-Intelligent Transportation Systems), it’s now known that they are an excellent mean for preventing road accidents, particularly those that directly assist the driver in the vehicle "ADAS - Advanced Driver Assistance System”. The introduction of ADAS systems on the market is slow to take, due to constraints such as cost and availability, as their availability is usually influenced by the commercial focus of car manufacturers, and is inherent in the use of high-end cars, or the purchase of costly extras. It is also considered that the technological requirements associated with these solutions are an obstacle to their implementation, since the average age of cars is high. The lack of reach of this type of solutions by the population, appears as motivation of this dissertation, for the development of a system capable of being accessible by all. This dissertation has, as main objective, the demonstration of an investigation in the area of the intelligent transport systems, being the base for the development of a prototype of an evolutive system of assistance to the driver, denominated Virtual Driver. Traffic signaling is the main mean of communication between the driver and the road, defining the rules and dangers inherent in their use. However, their non-visualization, by factors such as distraction, lack of focus on driving or lack of visibility, can compromise the safety of the driver himself and others present on the road, resulting in road accidents. Virtual Driver is a low cost solution that is executable through everyday devices such as smartphones or other micro-computers. In this way, it allows, in a first stage, to assist the driver through the recognition of vertical traffic signaling in real time, alerting him through audiovisual information. The developed prototype acts based on the Portuguese vertical traffic signalization, aiming later to cover other realities. In a final phase of this dissertation will be presented the evaluation of the performance of the solution, regarding the processes of detection and recognition of signaling, highlighting the use of the chosen methodology.
In today's society, road accidents remain one of the leading causes of death worldwide, killing millions of people each year and causing irreversible physical damage, resulting in total or partial disability. With the sustained growth of Intelligent Transportation Systems (ITS-Intelligent Transportation Systems), it’s now known that they are an excellent mean for preventing road accidents, particularly those that directly assist the driver in the vehicle "ADAS - Advanced Driver Assistance System”. The introduction of ADAS systems on the market is slow to take, due to constraints such as cost and availability, as their availability is usually influenced by the commercial focus of car manufacturers, and is inherent in the use of high-end cars, or the purchase of costly extras. It is also considered that the technological requirements associated with these solutions are an obstacle to their implementation, since the average age of cars is high. The lack of reach of this type of solutions by the population, appears as motivation of this dissertation, for the development of a system capable of being accessible by all. This dissertation has, as main objective, the demonstration of an investigation in the area of the intelligent transport systems, being the base for the development of a prototype of an evolutive system of assistance to the driver, denominated Virtual Driver. Traffic signaling is the main mean of communication between the driver and the road, defining the rules and dangers inherent in their use. However, their non-visualization, by factors such as distraction, lack of focus on driving or lack of visibility, can compromise the safety of the driver himself and others present on the road, resulting in road accidents. Virtual Driver is a low cost solution that is executable through everyday devices such as smartphones or other micro-computers. In this way, it allows, in a first stage, to assist the driver through the recognition of vertical traffic signaling in real time, alerting him through audiovisual information. The developed prototype acts based on the Portuguese vertical traffic signalization, aiming later to cover other realities. In a final phase of this dissertation will be presented the evaluation of the performance of the solution, regarding the processes of detection and recognition of signaling, highlighting the use of the chosen methodology.
Description
Keywords
Sistemas inteligentes de transporte Sistemas avançados de assistência ao condutor Reconhecimento de sinalização de trânsito vertical Intelligent transport systems Advanced driver assistance systems Vertical traffic sign recognition system