A Robust Registration Method for Autonomous Driving Pose Estimation in Urban Dynamic Environment Using LiDAR

Abstract

The registration of point clouds in urban environments faces problems such as dynamic vehicles and pedestrians, changeable road environments, and GPS inaccuracies. The state-of-the-art methodologies have usually combined the dynamic object tracking and/or static feature extraction data into a point cloud towards the solution of these problems. However, there is the occurrence of minor initial position errors due to these methodologies. In this paper, the authors propose a fast and robust registration method that exhibits no need for the detection of any dynamic and/or static objects. This proposed methodology may be able to adapt to higher initial errors. The initial steps of this methodology involved the optimization of the object segmentation under the application of a series of constraints. Based on this algorithm, a novel multi-layer nested RANSAC algorithmic framework is proposed to iteratively update the registration results. The robustness and efficiency of this algorithm is demonstrated on several high dynamic scenes of both short and long time intervals with varying initial offsets. A LiDAR odometry experiment was performed on the KITTI data set and our extracted urban data-set with a high dynamic urban road, and the average of the horizontal position errors was compared to the distance traveled that resulted in 0.45% and 0.55% respectively.

A Robust Laser-Inertial Odometry and Mapping Method for Large-Scale Highway Environments

“…c) Correction: Once the measurements δ y m t+1 ∈ R 6 are calculated, the full error state posterior δ x m t+1 ∈ R 6 and covariance Σ t+1 ∈ R 15×15 can be updated through Eq. (11)(12)(13). The K ∈ R 15×6 represents Kalman gain of the normal KF update.…”

ECPC-ICP: A 6D Vehicle Pose Estimation Method by Fusing the Roadside Lidar Point Cloud and Road Feature

“…The former has higher accuracy, and most of them need a dense point cloud to extract features for matching calculations. Researchers made breakthroughs in feature extraction [24][25][26][27], corresponding points matching [28][29][30][31][32][33], iterative calculations [34][35][36][37], and so on. On the contrary, model-free methods had fewer requirements for point cloud data, good generalization, but worse positioning accuracy.…”

4-D SLAM: An Efficient Dynamic Bayes Network-Based Approach for Dynamic Scene Understanding

“…François [18] used a Bayesian model to assess whether a point is static or dynamic. Wang [19] and Bahraini [20] used an improved RANSAC algorithm to track moving objects, but it may lose the ability to track static objects in scenes where moving objects are dense. Jiang [21] presented a LiDARcamera SLAM system, which used a sparse subspace clustering-based motion segmentation method to build a static map in a dynamic environment.…”