Gauss süreçleri bazlı üç boyutlu genişletilmiş hedef takibi için negatif bilgi birleştirmesi

Extended target tracking refers to the estimation of extent of a target as well as its position, kinematics, and orientation. In this thesis, we compare performances of Gaussian process based extended target tracking methods. Additionally, we propose a method that uses negative information fusion in three-dimensional point cloud data to enhance extent estimates of the target. Wide-ranging simulations are carried out to demonstrate the performance of the proposed algorithm. All simulations are carried out on a modular environment to be able to easily integrate different scenarios and compare the performances of the different methods. Results are obtained with both simulated and real data to attain better performance comparisons.