An Aircraft Tracking Method In Simulated Infrared Image Sequences Based On Regional Distribution

Sijie Wu,Saisai Niu,Kai Zhang,Jie Yan
DOI: https://doi.org/10.1007/978-3-319-92753-4_27
2018-01-01
Abstract:Based on regional distribution, we propose an aircraft tracking method that exploits both gray-level value distribution and structural distribution. After analyzing aircraft characteristics, we form the region proposals by clustering the peak values in the equivalent topographic map. Based on the region proposals formed with clustering algorithm, we quantify the spatial distribution to describe the geometric structure of the target. Then we calculate the gray-level value distribution and structural distribution to obtain the regional distribution descriptor to be used for characterizing the structural and texture information on the target. Finally, we perform experiments on various trackers to quantitatively evaluate the performance of our aircraft tracking method. The experimental results demonstrate that our aircraft tracking method outperforms several other state-of-the-art tracking methods in simulated infrared image sequences.
What problem does this paper attempt to address?