Multiobject Tracking for Thresholded Cell Measurements

Thomas Kropfreiter,Jason L. Williams,Florian Meyer

2024-06-02

Abstract:In many multiobject tracking applications, including radar and sonar tracking, after prefiltering the received signal, measurement data is typically structured in cells. The cells, e.g., represent different range and bearing values. However, conventional multiobject tracking methods use so-called point measurements. Point measurements are provided by a preprocessing stage that applies a threshold or detector and breaks up the cell's structure by converting cell indexes into, e.g., range and bearing measurements. We here propose a Bayesian multiobject tracking method that processes measurements that have been thresholded but are still cell-structured. We first derive a likelihood function that systematically incorporates an adjustable detection threshold which makes it possible to control the number of cell measurements. We then propose a Poisson Multi-Bernoulli (PMB) filter based on the likelihood function for cell measurements. Furthermore, we establish a link to the conventional point measurement model by deriving the likelihood function for point measurements with amplitude information (AM) and discuss the PMB filter that uses point measurements with AM. Our numerical results demonstrate the advantages of the proposed method that relies on thresholded cell measurements compared to the conventional multiobject tracking based on point measurements with and without AM.

Signal Processing

What problem does this paper attempt to address?

### Problems Addressed by the Paper

The paper aims to address the issue of handling thresholded cell measurement data in multi-object tracking (MOT). Traditional multi-object tracking methods typically use point measurement data, which are converted from cell structures through a thresholding or detection process in a preprocessing stage. However, this approach may lose important information from the original cell measurements. This paper proposes a Bayesian framework-based multi-object tracking method that can directly handle thresholded but still cell-structured measurement data.

Specifically, the main contributions of the paper include:

1. **Derivation of the Likelihood Function**: The authors derive a likelihood function that systematically incorporates an adjustable detection threshold, allowing control over the number of cell measurements.
2. **Poisson Multi-Bernoulli (PMB) Filter**: Based on the aforementioned likelihood function, a PMB filter suitable for cell measurements is proposed.
3. **Connection with Traditional Point Measurement Models**: The paper establishes a connection with traditional point measurement models, deriving the likelihood function for point measurements with amplitude information (AM) and discussing the use of PMB filters with point measurements and AM.
4. **Performance Evaluation**: Numerical experiments demonstrate the advantages of the proposed PMB filter in handling thresholded cell measurements compared to traditional PMB filters (with or without AM), particularly in terms of tracking accuracy and runtime.

### Summary

By introducing a new measurement model and the corresponding PMB filter, the paper addresses the issue of handling thresholded cell measurement data in multi-object tracking. This method not only retains the structural information of the original cell measurements but also controls the number of measurements by adjusting the detection threshold, thereby improving tracking performance.