Determination of aircraft aspect angle based on shape analysis

Pose information of the aircraft target embodies its kinematic characteristics. The pose can be estimated from mono-ocular image data, because of the close relationship between the imaging shape and the relative pose. Aiming at the problem of aspect angle identification, a new valid algorithm which was suitable for arbitrary viewing direction was presented in this paper. Distance-angle function signature of the aircraft silhouette was figured out, then preprocessing including median filtering and angle expanding was performed. At last the peak of the signature curve was detected by using the least squares curve fitting method. Based on the distribution characteristics of the head, wing and tail of the aircraft, the aspect angle of the target could be recognized via the logical relationship between the peaks' amplitude and angle. Experimental results show that the proposed algorithm is simple and has a more widely applicability than the traditional methods, and the identification probability can reach above 90%.