An accurate pose estimation system isdeterminant to obtain a good parameteridentification result. Vision-based measurementis an alternative method for UAV’s poseestimation. However, with limited field of viewand vision distance, the vision-basedmeasurement is more suitable for a relativelysmall area or object like indoor MAV.This work presents development of a visionbased measurement system using Dual CameraTechnique for 6- DOF pose estimation. Thesystem consists of two outside-in video cameraswith 30 fps of frame rate, four onboard markers,an image filtering and Direct LinearTransformation algorithm. The system issimulated using 3D software. A 2D pair imagesfrom both video cameras are then processed toreconstruct the position and orientation of theobject.A flight test scenario is virtually simulatedto obtain parameters of aerodynamic andmoment coefficient of a micro coaxialhelicopter. The results show that a vision basedmeasurements offers a relatively straightforward data processing to obtain the defined parameters.

Alternative methods,Direct linear transformation,Dual cameras,Image filtering,Indoor MAV,Pose estimation,Position and orientations,Vision-based measurements

Direct linear transformation,Dual camera technique,Indoor MAV,Parameter identification,Pose estimation