Aircraft tracking based on KLT feature tracker and image modeling
Khawar Ali, Shoab Ahmad Khan, Usman Akram · 2013
Tracking of aircraft in an image using the feature tracking algorithms faces significant challenges under conditions of severe aircraft rotation, highly cluttered background presence, arrival and collision with other aircrafts and sun, excessive noise, and varying lightening conditions due to weather changes. A robust and real-time framework has been provided in this paper for tracking aircraft in low resolution images using image modeling and feature tracking techniques. The main focus of the algorithm is on utilization of Kanade-Lucas-Tomasi (KLT) feature tracker and the manipulation of those features for modeling and tracking of the object (aircraft) in attention. The focus of object tracking is focused on both the feature tracking and image modeling by the manipulation of KLT features. The features of KLT algorithm are manipulated to extract only the features of the aircraft and an image model of the aircraft using histogram, mean, and standard deviation is created which is utilized in the consecutive frames to track movements and collisions. The algorithm has been tested on self defined dataset of 18000 frames and the results are presented with high accuracy and efficiency.