2026-10-13T04:25:56Z https://www.scidb.cn/oai

10.57760/sciencedb.13337 2023-11-17T17:29:53Z

2023-11-17 Labeled filtering dataset for roadside lidar doi:10.57760/sciencedb.13337 en This dataset is intended to assist in evaluating the filtering algorithm of roadside LiDAR.We used three sensors (VLP-16, RSM1, RS-Helios-32) for data collection at six different locations, Specifically, it includes three intersections and three road sections. There are four folders in this dataset：history_csv: This folder is used to store historical point cloud data used for background construction, and is exported through visualization software such as Velview. In addition, there is also .npy files in the folder, which corresponds to the .csv files and facilitates the use of Python to read data. Note: The .csv data of different sensors may be inconsistent.history_pcap: All raw historical data collected, specifically in. pcap format.labeled: The files in this folder are labeled point cloud data used to evaluate the performance of filtering algorithms. Each .pcd file containes 4 columns, x,y,z,label, where label==0 means background points and label==1 means traffic object points.viewer_software_and_user_document: The files here are softwares for viewing .pcap files and the user manual of the LiDAR used. roadside lidar; filtering; semantic segmentation Hui Liu PUBLIC https://creativecommons.org/licenses/by-nc-sa/4.0/ dataset Science Data Bank