The identification of aircraft without enough point cloud data

With the flights of civil aviation increase recently, the need for precise and reliable surveillance sensor has become far stronger than ever before. Many Surveillance sensors have been mounted at the airport, but the accuracy can’t meet the requirement on the apron. The LiDAR sensor was used to surveillance the apron with the development of the LiDAR sensor and computer vision algorithm. When the Lidar sensor only get the point cloud data of a small part of the aircraft. The point cloud data is not enough to identify the aircraft without 3D models. In this paper, a method is proposed to identify the aircraft without enough point cloud data. In the method, take the previous point cloud data of the aircraft as the model. Take the point cloud data with only a small part of the aircraft as the scene. Compute the Signature of Histograms of OrienTations descriptor of the model and the scene. And use the correspondence grouping algorithm based on Hough voting scheme in order to cluster the point set of the aircraft.