Abstract—Ensuring robust and safe 3D object detection in outdoor environments requires addressing the challenges posed by adverse conditions such as rain, fog, snow, and varying lighting. However, the scarcity of diverse, labeled training, and test data reflecting these conditions hinders progress in this area. Synthetic data augmentation offers a promising solution to bridge this gap. In this paper, we evaluate the effectiveness of an existing physical rain model for augmenting lidar-based datasets. First, we validate the model by comparing its results to those generated by state-of-the-art simulation software, AURELION. Next, we apply the rain model to augment the KITTI 3D object detection dataset with varying rain intensities and assess the impact on a lidar-based object detection framework. Our results demonstrate that the physical rain model produces outputs nearly identical to AURELION. Furthermore, the augmented data reveal a significant degradation in detection performance across all evaluated object classes under increasing rain intensities. Retraining the detection model with the augmented data set substantially improves its robustness, even under heavy rainfall. These findings highlight the potential of synthetic data augmentation for enhancing the resilience of lidar-based 3D object detection systems in adverse weather conditions.
 
Keywords—3D object detection, sensor robustness, lidar simulation, data augmentation, adverse weather, rain simulation, autonomous systems


Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).