园区自动驾驶多模态目标感知数据集

本数据集使用自行组装的阿克曼结构小车，在其上搭载工控机与激光雷达、相机等传感设备，随后在小车行驶过程中录制传感器数据。采集开始前启动激光雷达与相机的ROS驱动程序，并使用ROS录制传感器生成的数据流，所有数据被存储进一个包文件中。每次室外数据采集完成后再额外采集标定数据，作为本次采集的标定来源。

This dataset utilizes a self-assembled Ackermann-structured small vehicle fitted with an industrial personal computer (IPC), LiDAR, cameras and other sensing equipment. Sensor data was recorded while the vehicle was in motion. Before initiating data collection, the ROS drivers for the LiDAR and camera were launched, and the sensor-generated data streams were recorded via ROS, with all data stored in a rosbag file. Following the completion of each outdoor data acquisition session, supplementary calibration data was collected as the calibration source for that particular acquisition.

搜集汇总

背景与挑战

背景概述

该数据集是一个用于自动驾驶多模态目标感知的园区数据集，通过搭载激光雷达和相机的阿克曼结构小车采集，包含3.38GB的传感器数据流，适用于多场景多模态融合感知算法的验证。