Rv
R. van der Voort
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
2 records found
1
This thesis presents a comparative analysis of multi-target tracking algorithms for estimating the kinematic state and physical extent of maritime vessels from 4D millimetre-wave radar point clouds. The research addresses a gap in understanding the trade-offs between algorithmic complexity and motion model fidelity for inland maritime navigation.
The study evaluates three tracking paradigms, Tracking-by-Detection (TBD), the Probability Hypothesis Density (PHD) filter, and the Poisson Multi-Bernoulli Mixture (PMBM) filter, each tested with linear (Constant Velocity, Constant Acceleration) and non-linear (Constant Turn Rate) motion models. To represent vessel shape, three extended object models were implemented: the Random Matrix, Star-Convex Gaussian Process, and Principal Axis models.
Performance was evaluated using a mix of simulated datasets with perfect ground truth and collected real-world datasets. A hyperparameter optimisation process was conducted for every algorithm combination using the Optuna framework, with the objective of minimising the Generalised Optimal Sub-Pattern Assignment metric using a specified cost function.
The results yielded several key, and at times counterintuitive, findings. Firstly, increased algorithmic complexity did not yield superior performance. The simpler TBD and PHD frameworks consistently outperformed the theoretically optimal but computationally intensive PMBM filter, which was prone to generating unstable tracks and excessive false positives. Secondly, non-linear motion models offered no significant advantage over the simpler linear models. This could be attributed to the slow dynamics of maritime vessels, the high sensor update rate, which makes linear extrapolation sufficient, and the fact that measurement noise often dominates any subtle gains from a more precise motion model. Thirdly, for extent estimation, the Random Matrix model demonstrated the best balance of accuracy, stability, and computational efficiency. The more complex Gaussian Process and Principal Axis models struggled with stability and practicality.
The study concludes that the optimal balance is achieved by pairing simple linear motion models with the Random Matrix extent model, yielding accurate cardinality estimation, reliable state tracking, and efficient computation. The findings underscore that robustness and interpretability outweigh algorithmic sophistication when designing maritime tracking systems based on 4D radar. Ultimately, the thesis establishes a principled performance baseline and highlights the diminishing returns of complexity in this domain. ...
The study evaluates three tracking paradigms, Tracking-by-Detection (TBD), the Probability Hypothesis Density (PHD) filter, and the Poisson Multi-Bernoulli Mixture (PMBM) filter, each tested with linear (Constant Velocity, Constant Acceleration) and non-linear (Constant Turn Rate) motion models. To represent vessel shape, three extended object models were implemented: the Random Matrix, Star-Convex Gaussian Process, and Principal Axis models.
Performance was evaluated using a mix of simulated datasets with perfect ground truth and collected real-world datasets. A hyperparameter optimisation process was conducted for every algorithm combination using the Optuna framework, with the objective of minimising the Generalised Optimal Sub-Pattern Assignment metric using a specified cost function.
The results yielded several key, and at times counterintuitive, findings. Firstly, increased algorithmic complexity did not yield superior performance. The simpler TBD and PHD frameworks consistently outperformed the theoretically optimal but computationally intensive PMBM filter, which was prone to generating unstable tracks and excessive false positives. Secondly, non-linear motion models offered no significant advantage over the simpler linear models. This could be attributed to the slow dynamics of maritime vessels, the high sensor update rate, which makes linear extrapolation sufficient, and the fact that measurement noise often dominates any subtle gains from a more precise motion model. Thirdly, for extent estimation, the Random Matrix model demonstrated the best balance of accuracy, stability, and computational efficiency. The more complex Gaussian Process and Principal Axis models struggled with stability and practicality.
The study concludes that the optimal balance is achieved by pairing simple linear motion models with the Random Matrix extent model, yielding accurate cardinality estimation, reliable state tracking, and efficient computation. The findings underscore that robustness and interpretability outweigh algorithmic sophistication when designing maritime tracking systems based on 4D radar. Ultimately, the thesis establishes a principled performance baseline and highlights the diminishing returns of complexity in this domain. ...
This thesis presents a comparative analysis of multi-target tracking algorithms for estimating the kinematic state and physical extent of maritime vessels from 4D millimetre-wave radar point clouds. The research addresses a gap in understanding the trade-offs between algorithmic complexity and motion model fidelity for inland maritime navigation.
The study evaluates three tracking paradigms, Tracking-by-Detection (TBD), the Probability Hypothesis Density (PHD) filter, and the Poisson Multi-Bernoulli Mixture (PMBM) filter, each tested with linear (Constant Velocity, Constant Acceleration) and non-linear (Constant Turn Rate) motion models. To represent vessel shape, three extended object models were implemented: the Random Matrix, Star-Convex Gaussian Process, and Principal Axis models.
Performance was evaluated using a mix of simulated datasets with perfect ground truth and collected real-world datasets. A hyperparameter optimisation process was conducted for every algorithm combination using the Optuna framework, with the objective of minimising the Generalised Optimal Sub-Pattern Assignment metric using a specified cost function.
The results yielded several key, and at times counterintuitive, findings. Firstly, increased algorithmic complexity did not yield superior performance. The simpler TBD and PHD frameworks consistently outperformed the theoretically optimal but computationally intensive PMBM filter, which was prone to generating unstable tracks and excessive false positives. Secondly, non-linear motion models offered no significant advantage over the simpler linear models. This could be attributed to the slow dynamics of maritime vessels, the high sensor update rate, which makes linear extrapolation sufficient, and the fact that measurement noise often dominates any subtle gains from a more precise motion model. Thirdly, for extent estimation, the Random Matrix model demonstrated the best balance of accuracy, stability, and computational efficiency. The more complex Gaussian Process and Principal Axis models struggled with stability and practicality.
The study concludes that the optimal balance is achieved by pairing simple linear motion models with the Random Matrix extent model, yielding accurate cardinality estimation, reliable state tracking, and efficient computation. The findings underscore that robustness and interpretability outweigh algorithmic sophistication when designing maritime tracking systems based on 4D radar. Ultimately, the thesis establishes a principled performance baseline and highlights the diminishing returns of complexity in this domain.
The study evaluates three tracking paradigms, Tracking-by-Detection (TBD), the Probability Hypothesis Density (PHD) filter, and the Poisson Multi-Bernoulli Mixture (PMBM) filter, each tested with linear (Constant Velocity, Constant Acceleration) and non-linear (Constant Turn Rate) motion models. To represent vessel shape, three extended object models were implemented: the Random Matrix, Star-Convex Gaussian Process, and Principal Axis models.
Performance was evaluated using a mix of simulated datasets with perfect ground truth and collected real-world datasets. A hyperparameter optimisation process was conducted for every algorithm combination using the Optuna framework, with the objective of minimising the Generalised Optimal Sub-Pattern Assignment metric using a specified cost function.
The results yielded several key, and at times counterintuitive, findings. Firstly, increased algorithmic complexity did not yield superior performance. The simpler TBD and PHD frameworks consistently outperformed the theoretically optimal but computationally intensive PMBM filter, which was prone to generating unstable tracks and excessive false positives. Secondly, non-linear motion models offered no significant advantage over the simpler linear models. This could be attributed to the slow dynamics of maritime vessels, the high sensor update rate, which makes linear extrapolation sufficient, and the fact that measurement noise often dominates any subtle gains from a more precise motion model. Thirdly, for extent estimation, the Random Matrix model demonstrated the best balance of accuracy, stability, and computational efficiency. The more complex Gaussian Process and Principal Axis models struggled with stability and practicality.
The study concludes that the optimal balance is achieved by pairing simple linear motion models with the Random Matrix extent model, yielding accurate cardinality estimation, reliable state tracking, and efficient computation. The findings underscore that robustness and interpretability outweigh algorithmic sophistication when designing maritime tracking systems based on 4D radar. Ultimately, the thesis establishes a principled performance baseline and highlights the diminishing returns of complexity in this domain.
Bachelor thesis
(2023)
-
F.A. Voetter, M.C. Otte, H. Van Lierde, M.R. Schutte, M.J. Wijkhuizen, R.A. Milla-Koch, R. van der Voort, T. Laier, V.R. Gottumukkala, W.V. Cijsouw, P.C. Roling, S. Stroobants, R.K. van der Zwaard, P.M.A. de Jong, T. Mathijssen
Weather balloons have been used for decades as a means of measuring atmospheric properties. Balloons are launched twice a day from almost 900 locations worldwide, carrying measurement instruments called radiosondes. Along with data from ground-based sensors, aircraft, and remote-sensing satellites, the collected meteorological data is fed into numerical weather forecasts and climate models. The weather balloons typically burst at altitudes of around 33 km, after which the radiosonde falls back with a parachute but without any control. This leads to large amounts of waste scattered over land and sea. While the balloon material itself is fully biodegradable, the loss of radiosondes and their associated electronics can be an environmental threat. Furthermore, the loss of the filling gases such as helium and hydrogen also poses a threat to the environment – helium is a non-renewable resource and hydrogen indirectly causes warming effects when released in the upper atmosphere. These factors present the need to develop sustainable and reusable alternatives to current weather balloons. In response to this pressing need, the BRAVO family of gliders has been developed, which is short for “Balloon-Released Aerial Vehicle for weather Observation”. This report provides a comprehensive overview of the design process of creating these gliders. It introduces the detailed final design and outlines the logistical and operational procedures associated with their usage. Furthermore, the report encompasses meticulous analyses of cost and market factors, technical risks, and sustainability considerations. To ensure reliability, the gliders undergo thorough verification and validation processes. Finally, the report concludes with gained insights, recommendations, and directions for future work.F
...
Weather balloons have been used for decades as a means of measuring atmospheric properties. Balloons are launched twice a day from almost 900 locations worldwide, carrying measurement instruments called radiosondes. Along with data from ground-based sensors, aircraft, and remote-sensing satellites, the collected meteorological data is fed into numerical weather forecasts and climate models. The weather balloons typically burst at altitudes of around 33 km, after which the radiosonde falls back with a parachute but without any control. This leads to large amounts of waste scattered over land and sea. While the balloon material itself is fully biodegradable, the loss of radiosondes and their associated electronics can be an environmental threat. Furthermore, the loss of the filling gases such as helium and hydrogen also poses a threat to the environment – helium is a non-renewable resource and hydrogen indirectly causes warming effects when released in the upper atmosphere. These factors present the need to develop sustainable and reusable alternatives to current weather balloons. In response to this pressing need, the BRAVO family of gliders has been developed, which is short for “Balloon-Released Aerial Vehicle for weather Observation”. This report provides a comprehensive overview of the design process of creating these gliders. It introduces the detailed final design and outlines the logistical and operational procedures associated with their usage. Furthermore, the report encompasses meticulous analyses of cost and market factors, technical risks, and sustainability considerations. To ensure reliability, the gliders undergo thorough verification and validation processes. Finally, the report concludes with gained insights, recommendations, and directions for future work.F