Car Damage Detection (CarDD)
收藏资源简介:
CarDD是中国科学院合肥物质科学研究院创建的首个公开大型车辆损伤检测数据集,包含4000张高分辨率车辆损伤图像,涵盖6种损伤类别,共计超过9000个精细标注的实例。数据集通过严格的图像收集、筛选和标注过程构建,旨在支持车辆损伤的分类、检测、分割和显著性检测任务。CarDD的创建不仅推动了车辆损伤评估技术的发展,也为计算机视觉领域的研究提供了新的挑战和机遇,特别是在解决车辆损伤检测和分割中的复杂问题上。
CarDD is the first publicly available large-scale vehicle damage detection dataset developed by the Hefei Institutes of Physical Science, Chinese Academy of Sciences. It contains 4,000 high-resolution vehicle damage images, covering 6 damage categories with a total of over 9,000 finely annotated instances. The dataset is constructed through a rigorous process of image collection, screening and annotation, aiming to support tasks including vehicle damage classification, detection, segmentation and salient object detection. The creation of CarDD not only promotes the advancement of vehicle damage assessment technologies, but also provides new challenges and opportunities for computer vision research, particularly in addressing complex problems in vehicle damage detection and segmentation.




