VRAI (Vehicle Re-identification for Aerial Image)

VRAI 是用于基于无人机的智能应用的大规模车辆 ReID 数据集。该数据集由 13, 022 个车辆实例的 137, 613 张图像组成。每个车辆实例的图像由两个 DJI 消费级无人机在不同位置的摄像头捕获，具有各种视角和飞行高度（15m 到 80m）。

VRAI is a large-scale vehicle ReID dataset tailored for drone-enabled intelligent applications. This dataset comprises 137,613 images spanning 13,022 unique vehicle instances. Images for each vehicle instance were captured by cameras onboard two DJI consumer-grade drones positioned at distinct locations, with diverse viewing angles and flight altitudes ranging from 15 meters to 80 meters.

搜集汇总

背景与挑战

背景概述

VRAI是一个大规模车辆重识别数据集，包含13,022个车辆实例的137,613张图像，适用于无人机智能应用和计算机视觉研究。