Mestrado em Engenharia de Produção (FCT)
URI Permanente para esta coleção
Navegar
Navegando Mestrado em Engenharia de Produção (FCT) por Assunto "Computer vision"
Agora exibindo 1 - 1 de 1
Resultados por página
Opções de Ordenação
Item Detecção de dimensões de bagagens aeroportuárias a partir de dispositivos de baixo custo(Universidade Federal de Goiás, 2023-12-20) Silva, Vitor de Almeida; Bandeira, Michelle Carvalho Galvão da Silva Pinto; http://lattes.cnpq.br/8988187129672006; Roriz Junior, Marcos Paulino; http://lattes.cnpq.br/1356289387726731; Roriz Junior, Marcos Paulino; Rabelo, Diogo de Souza; Guterres, Marcelo XavierRecent studies indicate a significant increase in passenger boarding times, ranging from 22 to 40 minutes between 1990 and 2009. One contributing factor to this delay is the check-in process, primarily stemming from baggage dimension verification. Incorrect dimensions not only impact luggage storage but also result in additional costs for the customer. In such cases, passengers may be required to pay extra fees and return to the check-in line for corrections. To address this issue, companies are investing in self-bag drop systems, where passengers take responsibility for measuring their baggage. However, challenges arise due to the varied shapes of luggage and the complexities of handling multiple checked items. This study reveals that certain devices, particularly those utilizing LASER technology, can automatically obtain baggage dimensions. Despite their effectiveness, these devices often come with a significant price, and may require adjustments to the environment for installation. Consequently, this research aims to explore the technical feasibility of employing low-cost devices to measure airport luggage dimensions. Among the low-cost techniques, the use of the Microsoft Kinect depth sensor stands out, capable of obtaining a point cloud of the object under analysis. Based on this sensor, an algorithm was developed to capture, assemble and analyse the point cloud generated, thus obtaining the size of the luggage. To validate the approach, a prototype was built that contains a mat and a structure to fix the sensor, allowing the configuration of speed and data capture parameters, such as sampling step and capture region. The conducted tests indicate that the Microsoft Kinect V2 depth sensor can accurately capture depth, width, and height data. These results indicate the potential of this low-cost alternative in streamlining airline boarding operations.