Indoor hovering flight of a micro helicopter by markerless visual servo control

This paper proposes a vision-based localization and control algorithm for a micro helicopter with a wireless camera in a room. The 3D position, the posture and noise information are derived by a parallel template matching technique with edge detections and Gaussian filtering. The proposed algorithm achieves long-time hovering in a room. The flight experiment does not require any specified markers, texture information given in advance, or any special settings.