Extended target tracking using reduced rank gaussian processes

Download
2021-2-12
Özcan , Mustafa Buğra
Conventional tracking algorithms are predominantly based on point target assumption; however, this assumption is challenged as a result of the advents in sensor resolutions. Improvements on processors and rapid advances in sensor capabilities has enabled to the perception of target characteristics beyond the kinematics. Extended target tracking is the ability to learn target shapes that occupy multiple resolution cells and to track the motion of the target in a recursive framework. Gaussian process, a non-parametric method to settle the bridge between inputs and outputs of a system with tractable math, is opted for this thesis to establish the extended target tracking structure. On the other hand, computational complexity can be considered as a cost of the flexible nature of Gaussian processes. We, therefore, investigate the periodic kernel spectral approximation that relies on restating the kernel matrix with a lower rank. Following this, we perform in-depth mathematical derivations, conduct simulations in comparison with other approaches, and discuss several aspects of our extended target tracking method.

Suggestions

Extended Target Tracking using a Gaussian-Mixture PHD Filter
Granstrom, Karl; Lundquist, Christian; Orguner, Umut (2012-10-01)
This paper presents a Gaussian-mixture (GM) implementation of the probability hypothesis density (PHD) filter for tracking extended targets. The exact filter requires processing of all possible measurement set partitions, which is generally infeasible to implement. A method is proposed for limiting the number of considered partitions and possible alternatives are discussed. The implementation is used on simulated data and in experiments with real laser data, and the advantage of the filter is illustrated. S...
Extended Object Tracking and Shape Classification
Tuncer, Barkın; Kumru, Murat; Alatan, Abdullah Aydın; Özkan, Emre (2018-07-10)
Recent extended target tracking algorithms provide reliable shape estimates while tracking objects. The estimated extent of the objects can also be used for online classification. In this work, we propose to use a Bayesian classifier to identify different objects based on their contour estimates during tracking. The proposed method uses the uncertainty information provided by the estimation covariance of the tracker.
Surface Vessel Tracking in Airborne Infrared Imagery
Cakiroglu, Ahmet; Ulusoy, İlkay (2019-01-01)
Visual target tracking has been studied for decades and still remains a challenging problem. Ship tracking on infrared images has numerous challenges compared to conventional target tracking such as fast changing of appearance. Rapid appearance change caused by the manoeuvring movement of the target of image acquiring platform, confusion and occlusion caused by the active countermeasures employed by the target and disguise by cooling systems causes the target tracking algorithms to have low performance. In ...
Extended Target Tracking and Classification Using Neural Networks
Tuncer, Barkın; Kumru, Murat; Özkan, Emre (2019-01-01)
Extended target/object tracking (ETT) problem involves tracking objects which potentially generate multiple measurements at a single sensor scan. State-of-the-art ETT algorithms can efficiently exploit the available information in these measurements such that they can track the dynamic behaviour of objects and learn their shapes simultaneously. Once the shape estimate of an object is formed, it can naturally be utilized by high-level tasks such as classification of the object type. In this work, we propose ...
Fully-Automatic Target Detection and Tracking for Real-Time, Airborne Imaging Applications
Alkanat, Tunc; Tunali, Emre; Oz, Sinan (2015-03-14)
In this study, an efficient, robust algorithm for automatic target detection and tracking is introduced. Procedure starts with a detection phase. Proposed method uses two alternatives for the detection phase, namely maximally stable extremal regions detector and Canny edge detector. After detection, regions of interest are evaluated and eliminated according to their compactness and effective saliency. The detection process is repeated for a predetermined number of pyramid levels where each level processes a...
Citation Formats
M. B. Özcan, “Extended target tracking using reduced rank gaussian processes,” M.S. - Master of Science, Middle East Technical University, 2021.