BIDCD - Bosch Industrial Depth Completion Dataset
收藏资源简介:
BIDCD数据集是由博世人工智能中心和以色列理工学院合作创建的,专注于工业金属对象的RGBD数据集。该数据集通过安装在机器人操纵器上的深度相机收集,旨在促进特定领域深度完成模型的训练,应用于物流和制造任务。数据集包含约33,000张RGBD图像,覆盖了多种工业对象,如螺丝、圆柱体和散热器。创建过程中,使用了多视角融合技术来生成精确的深度图。该数据集的应用领域包括机器人操纵和自动化制造,旨在解决工业环境中深度感知和3D信息提取的问题。
The BIDCD dataset is an RGBD dataset focused on industrial metal objects, jointly created by Bosch Center for Artificial Intelligence and Technion – Israel Institute of Technology. This dataset was collected using a depth camera mounted on a robotic manipulator, and aims to facilitate the training of domain-specific depth completion models for logistics and manufacturing tasks. The dataset contains approximately 33,000 RGBD images, covering a variety of industrial objects such as screws, cylinders, and radiators. During its development, multi-view fusion technology was employed to generate accurate depth maps. The application scenarios of this dataset include robotic manipulation and automated manufacturing, and it is designed to address the challenges of depth perception and 3D information extraction in industrial environments.




