| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.29 MB | Adobe PDF |
Authors
Abstract(s)
Com a chegada da visão computacional, cada vez mais os investigadores estudam novos métodos para aquisição de imagens 3D de objetos para poderem ser úteis no nosso dia-a-dia na recolha de informações tanto nas áreas de engenharia, medicina, arquitetura, educação, entre outras. O lançamento do sensor Microsoft Kinect para a Xbox, projetado como acessório para detetar o movimento dos jogadores, veio chamar a atenção, não só pelo seu preço acessível, mas também por conseguir disponibilizar informações sem criptografia, permitindo que fosse usado para outros fins além dos videojogos. Este trabalho teve como base a análise e investigação da capacidade de obter informações acerca das distâncias em profundidade, deste sensor e, aplicá-la aos movimentos reativos efetuados pelo robô TurtleBot permitindo desta forma a interação com o meio e tornar a sua movimentação autónoma. O algoritmo utilizado em linguagem C++, na estrutura de software ROS (Robot Operating System ), é uma síntese do método de mudança de direção baseado no centro (centroid em inglês), dos pontos detetados nos objetos, que permitirá o desvio do robô perante um obstáculo e, um método com clustering de PCL (Point Cloud Library), que permitirá o reconhecimento de humanos.
With the arrival of computer vision, researchers have been doing research in methods to capture 3D images of objects, in order to use the information collected in different areas such as medicine, architecture, engineering, education, and among others. The launch of the Microsoft Kinect sensor for Xbox, designed as an accessory to detect the movement of the players came calling the attention of researchers, not only by your affordable price but also for getting available information unencrypted, allowing it to be used for other purposes than videogames. In this work, have been explored the ability to get information about the distances in depth, of this sensor, and apply it to the movements made by the TurtleBot robot, thus enabling the interaction with the environment and make their movement autonomously. The algorithm used in C language + +, on software structure of ROS (Robot Operating System ), is a summary of the method of changing direction based on the centroid from detected points on objects, which will allow the change of direction of the robot when it finds an obstacle, and a method, with clustering of PCL (Point Cloud Library), which will allow the recognition of humans.
With the arrival of computer vision, researchers have been doing research in methods to capture 3D images of objects, in order to use the information collected in different areas such as medicine, architecture, engineering, education, and among others. The launch of the Microsoft Kinect sensor for Xbox, designed as an accessory to detect the movement of the players came calling the attention of researchers, not only by your affordable price but also for getting available information unencrypted, allowing it to be used for other purposes than videogames. In this work, have been explored the ability to get information about the distances in depth, of this sensor, and apply it to the movements made by the TurtleBot robot, thus enabling the interaction with the environment and make their movement autonomously. The algorithm used in C language + +, on software structure of ROS (Robot Operating System ), is a summary of the method of changing direction based on the centroid from detected points on objects, which will allow the change of direction of the robot when it finds an obstacle, and a method, with clustering of PCL (Point Cloud Library), which will allow the recognition of humans.