A LOW-COST POSITION AND VELOCITY ESTIMATION SYSTEM FOR MULTIROTOR AERIAL VEHICLES USING A STATIC KINECT SENSOR
Luiz Manoel Santos Santana1; Davi Antônio dos Santos1; Ícaro Bezerra Viana1; Raphael Ballet1
1 Instituto Tecnológico de Aeronáutica
Resumo
Indoor estimation of position and velocity of a multirotor aerial vehicle (MAV) is frequently required for experimental evaluation of control laws and on-board navigation systems. The present work proposes a low-cost indoor position and velocity estimation system using a static Kinect sensor xed on the ceiling and pointing downwards. The Kinect sensor features an RGB camera, an infrared projector and a CMOS camera. Using the RGB camera, it is possible to determine the horizontal position of the vehicle. On the other hand, the infrared projector together with the CMOS camera provide depth images which allows altitude estimation. The Kinect measurements are acquired in a desktop computer and processed in MATLAB. The data processing includes 3D position determination from a camera model and 3D position/velocity estimation using a Kalman lter. The proposed system is demonstrated and evaluated experimentally.
Palavras-chave: State estimation; Multirotor aerial vehicles; Kinect sensor; Kalman filter