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.
Highlights
Pose estimation is one of the key technologies for autonomous driving in the urban environment.Considering the fact that the GPS is unable to keep high accuracy in urban environments, a very common solution to this problem has emerged in recent years
Part B is the LiDAR odometry experiment in the public dataset provided by KITTI [21] and a data set of urban highly dynamic environments collected by ourselves
The registration effect of method 3 has only decreased slightly while method 1 decreased greatly. Using both the improved segmentation and the improved RANSAC, the robustness of the registration is greatly improved in high dynamic environments with high initial error, which can be seen from the contrast between method 1
Summary
Pose estimation is one of the key technologies for autonomous driving in the urban environment. The most common solution is based on the pose provided by GPS/INS, where real time data acquired from LiDAR or camera are registered with a priori map to achieve accurate localization [1,2]. The second instance involves tall buildings, overpasses, and trees that may lead to multipathing and shadowing These attributes characterize the most important factors of GPS error, whereby should the precision of INS be not high enough, will lead to the easy generation of large pose error in order to make registration hard. A common coarse registration method is the RANSAC [5], which is mainly used to solve model estimation problems of large amounts of outliers in data sets.
Published Version (Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.