UDTIRI
收藏资源简介:
UDTIRI数据集由同济大学开发,专注于智能道路检测领域,包含1000张RGB图像及其像素/实例级标注,适用于多种光照和天气条件下的道路检测任务。数据集旨在通过提供大规模、高质量的标注数据,推动深度学习技术在城市数字双胞胎(UDT)中的应用,特别是在智能道路检测(IRI)领域。UDTIRI数据集支持对象检测、语义分割和实例分割网络的训练与评估,为解决道路维护和自动驾驶中的关键问题提供支持。
The UDTIRI Dataset is developed by Tongji University, focusing on the domain of intelligent road detection. It contains 1000 RGB images with pixel-level and instance-level annotations, and covers diverse illumination and weather conditions for road detection tasks. This dataset aims to advance the application of deep learning technologies in Urban Digital Twins (UDT), particularly in the field of Intelligent Road Detection (IRI), by providing large-scale, high-quality annotated data. The UDTIRI dataset supports the training and evaluation of object detection, semantic segmentation, and instance segmentation networks, and provides support for addressing key challenges in road maintenance and autonomous driving.




