Safe and efficient airport operations are highly dependent on the availability of reliable visual scene information. Especially at airports without A-SMGCS support, the controller has to combine several input sources to interpret the overall situation on the ground. LiDAR sensor technology combined with computer vision algorithms for object detection and tracking was identified as a cost-effective method to support surveillance tasks of the controller, especially with regard to non-cooperative objects. Following previous research on semantic segmentation of airport scenes, this paper deals with a real-time detection and tracking method of fast-moving objects on the apron in labeled LiDAR scans. The method integrates a novel point sampling strategy into a basic Kalman filter by selecting a subset of representative points from the scanned points as a function of the velocity and distance of a moving object from the sensor. We show that the accuracy of the proposed detection and tracking method is in-line or close to ICAO A-SMGCS standards up to a distance of 350 m, covering a typical small apron.