Superelliptical extended target tracking with contour measurements

Date

2024-8-16

Author

Yurdakul, Oğul Can

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

108
views

0
downloads

Cite This

In this thesis, we propose a measurement model using a family of curves called Lamé curves or superellipses for extended target tracking problems with contour measurements. This family of curves contains pinched or regular rhombuses, ellipses, and rounded rectangles, all available through changing a single scalar parameter we call the superellipse exponent. Therefore, the proposed measurement model can express the target extent with minimal parameters, including this superellipse exponent and the centroid coordinates, half-lengths and orientation angle. Such a measurement model is expressive enough for many objects of interest, especially in urban scenarios, while being computationally lightweight for applications where resources may be limited. We propose two inference methods using this newly proposed extent model: A Rao-Blackwellized particle filter with an implicit measurement equation and an iterated extended Kalman filter with an explicit measurement equation. Additionally, we derive simple analytical conditions to tackle the self-occlusion problem, allowing our proposed inference methods to update an extension estimate only when the model deems it is visible. Our simulation and real data studies demonstrate that the proposed methods are computationally efficient and outperform various extended target tracking algorithms with contour measurements.

Subject Keywords

Extended target tracking, Contour measurements, Sensor-object geometry, Superellipse, Lamé curve

URI

https://hdl.handle.net/11511/111009

Collections

Graduate School of Natural and Applied Sciences, Thesis