2017年基于3D视觉的花椰菜头部检测、定位和尺寸确定数据集

三维西兰花数据集为基于三维视觉的西兰花头状花序检测而创建，旨在进行选择性机器人收割。图像数据使用安装在拖拉机上的红绿蓝-深度相机（Microsoft Kinect v2）在人工照明条件下，在不同西兰花田间采集。该数据集由16个录制视频和随附三维点云数据文件，以及一组大小为1920×1080像素的png格式彩色图像组成。该数据集可为使用三维点云和彩色图像评估对象分割和定位任务提供有效数据资源。https://lcas.lincoln.ac.uk/nextcloud/shared/agritech-datasets/broccoli/broccoli_datasets.htmlThe 3D Broccoli Dataset was created for 3D vision-based broccoli head detection, aiming at selective robotic harvesting. Image data was collected in various broccoli fields under artificial lighting conditions using an RGB-D camera (Microsoft Kinect v2) mounted on a tractor. This dataset consists of 16 recorded videos and their accompanying 3D point cloud data files, as well as a set of PNG-format color images with a resolution of 1920×1080 pixels. This dataset serves as an effective data resource for evaluating object segmentation and localization tasks using 3D point clouds and color images. https://lcas.lincoln.ac.uk/nextcloud/shared/agritech-datasets/broccoli/broccoli_datasets.html