Performance Evaluation of the Highway Radar Occupancy Grid

Abstract

The quality of environmental perception is crucial for automated vehicle capabilities. In order to ensure the required accuracy, the occupancy grid mapping algorithm is often utilised to fuse data from multiple sensors. This paper focuses on the radar-based occupancy grid for highway applications and describes how to measure effectively the quality of the occupancy map. The evaluation was performed using the novel grid pole-like object analysis method. The proposed assessment is versatile and can be applied without detailed ground truth information. The evaluation was tested with a simulation and real vehicle experiments on the highway.

Sensor Visibility Estimation: Metrics and Methods for Systematic Performance Evaluation and Improvement

“…Porębski and Kogut [16] propose an evaluation method for occupancy grids by assessing the correctness of the shapes of landmarks such as road signs.…”

“…• Occupancy: A 2D Cartesian occupancy grid with resolution 1m x 1m is processed with the filtered data: The inverse sensor model p(Occ m |z) estimates the occupancy of m given measurement z by a scaled multivariate Gaussian probability density function (PDF) as in Elfes [23]. A decay function accounts for dynamic objects as in Porębski and Kogut [16]. • Visibility: V m is calculated based on Occ m by the lineof-sight approach similar to Adarve et al [24]: Cells between the sensor origin and last occupied cell are marked visible, cells behind invisible.…”

Situation-Aware Environment Perception Using a Multi-Layer Attention Map

“…OSPA [33] or (generalized) IoU [34] and F 1 -Score for object detection. As another example, Porębski and Kogurt [35] introduced a metric for occupancy grid evaluation. These metrics compare the performance of a module under test with the ground truth.…”