2020/11/06 by Keenan Burnett, Angela P. Schoellig, Burnett, Keenan +3
Engineering · Environmental Science · Physics and Astronomy · #Robotics and Sensor-Based Localization #Remote Sensing and LiDAR Applications #Advanced Optical Sensing Technologies
paper · pdf · doi:10.48550/arxiv.2011.03512
In order to tackle the challenge of unfavorable weather conditions such as\nrain and snow, radar is being revisited as a parallel sensing modality to\nvision and lidar. Recent works have made tremendous progress in applying\nspinning radar to odometry and place recognition. However, these works have so\nfar ignored the impact of motion distortion and Doppler effects on\nspinning-radar-based navigation, which may be significant in the self-driving\ncar domain where speeds can be high. In this work, we demonstrate the effect of\nthese distortions on radar odometry using the Oxford Radar RobotCar Dataset and\nmetric localization using our own data-taking platform. We revisit a\nlightweight estimator that can recover the motion between a pair of radar scans\nwhile accounting for both effects. Our conclusion is that both motion\ndistortion and the Doppler effect are significant in different aspects of\nspinning radar navigation, with the former more prominent than the latter. Code\nfor this project can be found at:\nhttps://github.com/keenan-burnett/yetiradarodometry\n