Calibration and Stability Analysis of the VLP-16 Laser Scanner

We report on a calibration and stability analysis of the Velodyne VLP-16 LiDAR scanner. The sensor is evaluated for long-term stability, geometric calibration and the effect of temperature variations. To generalize the results, three separate VLP-16 sensors were examined. The results and conclusions from the analysis of each of the individual sensors was similar. We found that the VLP-16 showed a consistent level of performance, in terms of range bias and noise level over the tested temperature range from 0–40 °C. A geometric calibration was able to marginally improve the accuracy of the VLP-16 point cloud (by approximately 20%) for a single collection, however the temporal stability of the geometric calibration negated this accuracy improvement. Overall, it was found that there is some long-term walk in the ranging observations from individual lasers within the VLP-16, which likely causes the instability in the determination of geometric calibration parameters. However, despite this range walk, the point cloud delivered from the VLP-16 sensors tested showed an accuracy level within the manufacturer specifications of 3 cm RMSE, with an overall estimated RMSE of range residuals between 22 mm and 27 mm.