KITTI-road

The road and lane estimation benchmark consists of 289 training and 290 test images. We evaluate road and lane estimation performance in the bird's-eye-view space. It contains different categories of road scenes: We equipped a standard station wagon with two high-resolution color and grayscale video cameras. Accurate ground truth is provided by a Velodyne laser scanner and a GPS localization system. Our datsets are captured by driving around the mid-size city of Karlsruhe, in rural areas and on highways. Up to 15 cars and 30 pedestrians are visible per image.

该道路与车道估计基准数据集共包含289张训练图像与290张测试图像。我们将在鸟瞰视图（bird's-eye-view）空间内评估道路与车道估计模型的性能。该数据集涵盖多类道路场景：我们采用一辆标准旅行车作为搭载平台，其上安装了两台高分辨率彩色与灰度摄像机。数据集的精准真值标注（ground truth）由Velodyne激光扫描仪与GPS定位系统提供。本数据集采集自德国卡尔斯鲁厄中型城市、乡村区域以及高速公路沿线的行车场景。单张图像中最多可同时出现15辆汽车与30名行人。

搜集汇总

背景与挑战

背景概述

KITTI-road是一个用于道路和车道线估计的2D语义分割数据集，包含579张图像（289训练/290测试），涵盖多种城市道路场景。数据通过高分辨率摄像头、激光扫描仪和GPS系统采集，包含城市、乡村和高速公路等复杂交通场景。