RGB-D Frames Dataset
收藏资源简介:
该数据集由捷克技术大学的研究团队创建,旨在比较四种立体深度感知相机(Intel RealSense D435、D455、StereoLabs ZED 2和Luxonis OAK-D Pro)在机器人应用中的性能。数据集包含超过12,000帧RGB-D图像,涵盖了平面表面感知、塑料娃娃感知和YCB数据集中的家庭物品感知三种场景。数据集的创建过程包括在不同距离下记录静态场景,并通过点云分割和地面真值对比来评估相机性能。该数据集可用于机器人视觉任务中的深度感知、物体分割和抓取等应用,旨在为机器人社区提供可靠的深度感知数据基准。
This dataset was created by a research team at the Czech Technical University, aiming to compare the performance of four stereo depth perception cameras (Intel RealSense D435, D455, StereoLabs ZED 2, and Luxonis OAK-D Pro) in robotic applications. It contains over 12,000 RGB-D image frames, covering three scenarios: planar surface perception, plastic doll perception, and household object perception from the YCB dataset. The dataset creation process involved recording static scenes at varying distances, and evaluating camera performance through point cloud segmentation and ground truth comparison. This dataset can be applied to tasks such as depth perception, object segmentation, and grasping in robotic vision, and aims to provide a reliable benchmark for depth perception data within the robotics community.




