Variable motion model for lidar motion distortion correction

Cameras capture every pixel in transient duration for one frame image, while lidars capture every point in relevant long duration for one frame point cloud. When lidars stay static, point cloud will not cause motion distortion, when lidars move, point cloud will have motion distortion, this issue becomes worse with the increase of lidars’ linear velocity and angular velocity. As an inertial space movements’ sensor, IMU (inertial measurement unit) captures motions by a high frequency, which can be utilized to correct lidars motion distortion. In this paper, a variable motion model to correct lidars motion distortion is proposed. Comparing to the constant motion model, the proposed model has better performance on handling distortion caused by varying motion. Experiments results show that the variable motion model has a beneficial influence on distortion correction.