PHYSICS-BASED LIDAR WAVEFORM SIMULATION METHOD FOR REALISM IMPROVEMENT OF DRIVING SIMULATORS
- Delivery
- Available on this site
- Format
- Price
- Non-members (tax incl.)：￥1,100 Members (tax incl.)：￥880
- Publication code
- 20219050
- Paper/Info type
- Other International Conferences
- Pages
- 1-6(Total 6 p)
- Date of publication
- Sep 2021
- Publisher
- JSAE
- Language
- English
Detailed Information
| Author(E) | 1) Quang Nhat Nguyen, 2) Alexander Carballo, 3) Kazuya Takeda |
|---|---|
| Affiliation(E) | 1) Nagoya University, 2) Nagoya University;TierIV Inc., 3) Nagoya University;TierIV Inc. |
| Abstract(E) | Driving simulators are crucial to the development of automobiles, especially autonomous ones. They allow unlimited and iterative testing in any driving conditions, including those that are too adverse to be possible for testing in the real world, thus help to enhance the safety and stability of the autonomous driving system (ADS) in development. However, currently, simulation results may be compromised if LIDAR simulation is involved because as far as our knowledge goes, there is no driving simulator capable of simulating LIDAR accurately according to the rules of Physics. Since many ADS’s rely heavily on LIDAR signal, this leads to the need for additional real-world testing that is expensive, risky, and inconvenient. Thus, it is important to have an accurate simulation of LIDAR. Our approach is inspired by the core principle of LIDAR’s operation: the signal processing of LIDAR waveforms, which are generated by the photosensors’ signals after being handled by the electrical and logical components. Our work proposes a photon-tracing method for LIDAR waveform simulation based on tracing the Physics interaction of photons with the environment. We also include a comparison of the simulated waveforms with the actual LIDAR waveforms recorded in the same conditions. Evaluation results verify that our method behaves consistently with the rules of Physics and confirm our research purpose. Our proposed method sets a step towards highly realistic driving simulators and provides an efficient tool for the research and development of LIDAR sensors. |
