Aerial Target Detection and Tracking in Infrared Image Sequences by using Morphological Operations, Kalman Filtering and Elliptical Representation

In this paper, a novel algorithm for aerial target detection and tracking in infrared image sequences is proposed. In the detection step, firstly the dint target is enhanced by applying Laplacian of Gaussian (LoG) filter and modified white top-hat transformation. Then the true targets are extracted by implementing thresholding segmentation procedures. In the tracking step, the next position of the target is predicted by the Kalman filter and the detection algorithm is applied in the neighborhood surrounding the predicted position. To better represent the target, the moment features are used and the target is covered by a suitable ellipse. The evaluated performances demonstrate that suggested algorithms are more efficient compared with other commonly used detection and tracking methods.