RoRaTrack
收藏资源简介:
RoRaTrack数据集是由普渡大学电子与计算机工程学院的研究团队创建的,专为自动驾驶赛车道路场景设计的开源数据集。该数据集包含来自赛车上的多角度相机的图像数据和实例级别的注释,如分割掩码。数据是在印第安纳州的一条赛车轨道上,使用Dallara AV-21赛车收集的。RoRaTrack旨在解决高速模糊、颜色反转、轨道上缺乏车道标记等常见问题。数据集分为正常、曲线道路、颜色失衡、模糊和眩光等多种类别,包含两个相机角度的607张左前图像和791张右前图像,总计1398张图像,分为训练和测试集。该数据集适用于开发和研究自动驾驶赛车中的轨道检测算法。
RoRaTrack is an open-source dataset specifically designed for autonomous racing road scenarios, created by a research team from the School of Electrical and Computer Engineering at Purdue University. It includes image data from multi-angle cameras mounted on the racing car and instance-level annotations such as segmentation masks. The dataset was collected on a race track in Indiana using a Dallara AV-21 racing car. RoRaTrack aims to address common challenges in autonomous racing scenarios, including high-speed motion blur, color inversion, and the lack of lane markings on the track. The dataset is categorized into multiple types such as normal, curved road, color imbalance, blur, and glare. It contains 607 front-left images and 791 front-right images from two camera angles, totaling 1398 images, which are split into training and test sets. This dataset is suitable for developing and researching track detection algorithms in autonomous racing.




