Wheat 3D PartNet: Annotated Dataset for Part segmentation of wheat

Wheat3D PartNet is a large-scale 3D point cloud dataset for wheat phenotyping, containing 1,300+ annotated samples of three cultivars (Paragon, Gladius, Apogee). Each point cloud is manually segmented into ear and non-ear regions, enabling tasks such as 3D plant part segmentation, spike counting, and trait measurement. The dataset provides benchmark train/test splits and is openly available for research in deep learning, phenotyping, and precision agriculture.Corresponding author contact: [e-mail]