毫米波雷达与摄像头感知数据

无人机视角的毫米波雷达与摄像头感知数据：项目组在无人机上部署摄像头与毫米波雷达，并感知环境中目标的位置，从而生成多模态感知数据。该数据集包含正常天气、夜晚和雾天三种场景下的图像和毫米波雷达点云数据。数据集的收集时间为2021.10-2022.01，通过对摄像头和毫米波雷达进行时空对齐来标注数据，无人机飞行高度为8-10米，采集设备距离目标的距离为15-20米，毫米波雷达和摄像头的采集速率为20fps，无人机与边缘设备采用WiFi6路由器进行通信。无人机视角的毫米波雷达与摄像头感知数据集:数据体量455MB.恶劣条件下毫米波雷达与摄像头感知数据集：数据体量122MB

Drone-view Millimeter-wave Radar and Camera Perception Dataset: The project team deployed cameras and millimeter-wave radars on drones to perceive target positions in the surrounding environment, generating multimodal perception data. This dataset contains image and millimeter-wave radar point cloud data across three scenarios: normal weather, nighttime, and foggy conditions. The dataset was collected between October 2021 and January 2022, with all data annotated via spatiotemporal alignment of the camera and millimeter-wave radar. During data collection, the drone flew at an altitude of 8–10 meters, and the acquisition equipment was positioned 15–20 meters away from the targets. Both the millimeter-wave radar and camera operated at an acquisition rate of 20 fps, and the drone communicated with the edge device via a WiFi6 router. The Drone-view Millimeter-wave Radar and Camera Perception Dataset has a total data volume of 455 MB, while the Adverse Condition Millimeter-wave Radar and Camera Perception Dataset has a total data volume of 122 MB.

搜集汇总

背景与挑战

背景概述

该数据集是北京邮电大学在2021年10月至2022年1月期间，通过无人机搭载摄像头和毫米波雷达采集的多模态感知数据，包含正常天气、夜晚和雾天三种场景下的图像和点云数据，总数据量为510.22MB。数据集具有时空对齐标注，采集高度为8-10米、距离为15-20米，适用于人工智能和计算机科学技术领域，特别是恶劣天气环境下的感知算法研究。