CableInspect-AD
收藏资源简介:
CableInspect-AD是由加拿大公共事业公司Hydro-Québec的领域专家创建和标注的高质量公开数据集,专门用于机器人电力线检查中的视觉异常检测(VAD)。该数据集包含4798张高分辨率图像,涵盖三种电力线电缆上的6023个标注异常,代表了Hydro-Québec列出的七种最常见缺陷类型,具有不同的严重程度。数据集的创建过程包括使用移动摄像机捕获图像,并进行详细的图像级、像素级和边界框标注,以提供异常的详细分类。该数据集旨在解决现有数据集在捕捉真实世界多样化和意外异常方面的不足,推动VAD技术在复杂真实世界领域的应用。
CableInspect-AD is a high-quality public dataset created and annotated by domain experts from Hydro-Québec, a Canadian public utility company, specifically for visual anomaly detection (VAD) in robotic power line inspection. It contains 4,798 high-resolution images, covering 6,023 annotated anomalies across three types of power line cables, which represent the seven most common defect types listed by Hydro-Québec with varying severity levels. The dataset was developed by capturing images using mobile cameras, followed by detailed image-level, pixel-level, and bounding box annotations to provide detailed classification of the anomalies. This dataset aims to address the limitations of existing datasets in capturing diverse and unexpected real-world anomalies, so as to promote the application of VAD technologies in complex real-world domains.




