Robust sensing and perception in adverse weather conditions remain one of the biggest challenges for realizing reliable autonomous vehicle mobility services. Prior work has established that rainfall rate is a useful measure for the adversity of atmospheric weather conditions. This work presents a probabilistic hierarchical Bayesian model that infers rainfall rate from automotive lidar point cloud sequences with high accuracy and reliability. The model is a hierarchical mixture of experts model, or a probabilistic decision tree, with gating and expert nodes consisting of variational logistic and linear regression models. Experimental data used to train and evaluate the model is collected in a large-scale rainfall experiment facility from both stationary and moving vehicle platforms. The results show prediction accuracy comparable to the measurement resolution of a disdrometer, and the soundness and usefulness of the uncertainty estimation. The model achieves RMSE 2.42mm/h after filtering out uncertain predictions. The error is comparable to the mean rainfall rate change of 3.5mm/h between measurements. Model parameter studies show how predictive performance changes with tree depth, sampling duration, and crop box dimension. A second experiment demonstrates the predictability of higher rainfall above 300mm/h using a different lidar sensor, demonstrating sensor independence.
Probabilistic Rainfall Estimation from Automotive Lidar
05.06.2022
2415674 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Springer Verlag | 2015
|A Stochastic Physical Simulation Framework to Quantify the Effect of Rainfall on Automotive Lidar
SAE Technical Papers | 2019
|A Stochastic Physical Simulation Framework to Quantify the Effect of Rainfall on Automotive Lidar
British Library Conference Proceedings | 2019
|Automotive LiDAR Technology: A Survey
IEEE | 2022
|Robust 3D IMU-LIDAR calibration and multi sensor probabilistic state estimation
BASE | 2020
|