No description available.

In recent years wearable devices have been advancing at a rapid pace and one of the largest growing segments is the smart glass segment. In this thesis the feasibility of today’s ARM-based smart glasses are evaluated for automatic license plate recognition (ALPR). The license plate is by far the most prominent visual feature to identify a spe- cific vehicle, and exists on both old and newly produced vehicles. This thesis propose an ALPR system based on a sequence of vertical edge detection, a cascade classifier, verti- cal and horizontal projection as well as a general purpose optical character recognition library. The study further concludes that the optimal input resolution for license plate detection using vertical edges is 640x360 pixels and that the license plate need to be at least 20 pixels high or the characters 15 pixels high in order to successfully segment the plate and recognize each character. The separate stages were successfully implemented into a complete ALPR system that achieved 79.5% success rate while processing roughly 3 frames per second when running on a pair of Google Glass. / Under de senaste åren har området wearables avancerat i snabb takt, och ett av de snabbast växande segmenten är smarta glaögon. I denna examensuppsats utvärderas lämpligheten av dagens ARM-baserade smarta glasögon med avseende på automatisk registreringsskyltigenkänning. Registreringsskylten är den i särklass mest framträdande visuella egenskapen som kan användas för att identifiera ett specifikt fordon, och den finns på både gamla och nyproducerade fordon. Detta examensarbete föreslår ett system för automatisk registreringsskyltigenkänning baserat på en följd av vertikal kantdetektering, en kaskad av boostade klassificerare, vertikal och horisontell projektion samt ett optiskt teckenigenkänningsbibliotek. Studien konstaterar vidare att den optimala upplösningen för registreringsskyltdetektion med hjälp av vertikala kanter på smarta glasögonär 640x360 pixlar och att registreringsskylten måste vara minst 20 pixlar hög eller tecknen 15 pixlar höga för att registreringsskylten framgångsrikt skall kunna segmenteras samt tecken identifieras. De separata stegen implementerades framgångsrikt till ett system för automatisk registreringsskyltigenkänning på ett par Google Glass och lyckades känna igen 79,5% av de testade registreringsskyltarna, med en hastighet av ungefär 3 bilder per sekund.

This thesis evaluates the performance of a vehicle detection technology, Automated License Plate Recognition (ALPR) camera systems, with regards to its ability to produce real-time travel time information in active work zones. A literature review was conducted to investigate the ALPR technology as well as to identify other research that has been conducted using ALPR systems to collect travel time information. Next, the ALPR technology was tested in a series of field deployments in both an arterial and a freeway environment. The goal of the arterial field deployment was to evaluate the optimal ALPR camera angles that produce the highest license plate detection rates and accuracy percentages. Next, a series of freeway deployments were conducted on corridors of I-285 in Atlanta, Georgia in order to evaluate the ALPR system in active work zone environments. During the series of I-285 freeway deployments, ALPR data was collected in conjunction with data from Bluetooth and radar technologies, as well as from high definition video cameras. The data collected during the I-285 deployments was analyzed to determine the ALPR vehicle detection rates. Additionally, a script was written to match the ALPR reads across two data collection stations to determine the ALPR travel times through the corridors. The ALPR travel time data was compared with the travel time data produced by the Bluetooth and video cameras with a particular focus on identifying travel time biases associated with each given technology. Finally, based on the knowledge gained, recommendations for larger-scale ALPR work zone deployments as well as suggestions for future research are provided.

[pt] Sistemas de reconhecimento automático de placas (ALPR na sigla em inglês) são geralmente utilizados em aplicações como controle de tráfego, estacionamento, monitoração de faixas exclusivas entre outras aplicações. A estrutura básica de um sistema ALPR pode ser dividida em quatro etapas principais: aquisição da imagem, localização da placa em uma foto ou frame de vídeo; segmentação dos caracteres que compõe a placa; e reconhecimento destes caracteres. Neste trabalho focamos somente na etapa de reconhecimento. Para esta tarefa, utilizamos um Perceptron multiclasse, aprimorado pela técnica de geração de atributos baseada em entropia. Mostramos que é possível atingir resultados comparáveis com o estado da arte, com uma arquitetura leve e que permite aprendizado contínuo mesmo em equipamentos com baixo poder de processamento, tais como dispositivos móveis. / [en] ALPR systems are commonly used in applications such as traffic control, parking ticketing, exclusive lane monitoring and others. The basic structure of an ALPR system can be divided in four major steps: image acquisition, license plate localization in a picture or movie frame; character segmentation; and character recognition. In this work we ll focus solely on the recognition step. For this task, we used a multiclass Perceptron, enhanced by an entropy guided feature generation technique. We ll show that it s possible to achieve results on par with the state of the art solution, with a lightweight architecture that allows continuous learning, even on low processing power machines, such as mobile devices.