TimberVision
收藏资源简介:
TimberVision数据集由奥地利技术研究院创建,旨在为林业操作中的自动化任务提供支持。该数据集包含2023张RGB图像,标注了51,338个树干组件,涵盖了切割和侧面表面。数据来源包括森林、装载和收获场景,图像分辨率从1280x720到3000x2000像素不等。数据集通过半自动化的标注流程生成,适用于目标检测、实例分割和多目标跟踪等任务,旨在提高林业操作的效率和安全性。
Created by the Austrian Institute of Technology, the TimberVision dataset is designed to support automated tasks in forestry operations. This dataset includes 2,023 RGB images, with 51,338 annotated trunk components covering both cut surfaces and lateral surfaces. Data sources cover forest, loading and harvesting scenarios, and image resolutions range from 1280×720 to 3000×2000 pixels. Generated via a semi-automated annotation pipeline, the dataset is suitable for tasks such as object detection, instance segmentation and multi-object tracking, aiming to enhance the efficiency and safety of forestry operations.




