Abstract
<div class="section abstract"><div class="htmlview paragraph">Autonomous vehicles rely upon accurate information about their surrounding environment to perform safe operational planning. The environment sense and perception system normally produces camera image data and LiDAR point cloud data that are processed and then fused to obtain a better perception of the environment than is possible from either alone. The accuracy of the fused data depends upon knowledge of the position of each sensor on the ego vehicle. Vehicle damage, improper sensor installation, sensor mount deformation, mount movement excited by vehicle motion, and/or other situations can result in an unexpected position of the sensor. This error adds uncertainty into the sensor measurement fusion that is normally not accounted for. LiDAR translational offset and angular orientation misalignment errors are investigated for correction. The correction approach, derived from camera image information, is based upon an error-in-variables recursion that takes into account errors from LiDAR angular and ranging uncertainties that increase with distance due to beam spread. The proposed approach incorporates approximations of these errors to make continuous updates to the estimates of the LiDAR misalignment. The approach, with and without noise included, is compared to recursive least squares and nonlinear least squares in simulated traffic of various speeds and positions relative to the ego vehicle. Results show that nonlinear least squares provides the best overall performance (based upon an error metric) since it utilizes the whole data set at once to produce an estimate. However, the proposed approach without noise gave similar or better misalignment estimates compared to the other recursive methods and is the overall preferred recursive method since it has lower computational complexity than the proposed approach with noise. Further, recursive least squares provides acceptable online misalignment estimates at low computational complexity when the angular error is less than or equal to 10°.</div></div>