Data underlying the publication: 3D pose of tomato peduncle nodes using deep keypoint detection and point cloud

This is the dataset used for paper '3D pose of tomato peduncle nodes using deep keypoint detection and point cloud'. The dataset contains 503 train images and 149 test images and their depth images and pose annotations. To test the impact of different view poses for detection results. The images were taken from 5 different angles for each peduncle node.

本数据集为论文《基于深度关键点检测与点云的番茄果柄节点三维位姿》所使用的数据集。本数据集包含503张训练图像、149张测试图像，以及对应的深度图像与位姿标注信息。为探究不同拍摄视角对检测结果的影响，本数据集针对每个果柄节点，采用5种不同角度采集图像。