MMCBE
收藏资源简介:
MMCBE数据集,由澳大利亚联邦科学与工业研究组织创建,是首个多模态作物生物量估算数据集。该数据集包含216组多视角无人机图像和LiDAR点云,以及手工标注的地面真实数据。创建过程中,数据收集自澳大利亚的干旱棉花田,通过无人机和LiDAR技术进行。MMCBE数据集主要应用于作物生物量预测,同时探索了3D作物重建和新型视角合成等计算机视觉任务,旨在解决高分辨率作物生物量估算的挑战。
The MMCBE dataset, created by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) of Australia, is the first multimodal crop biomass estimation dataset. It comprises 216 sets of multi-view UAV images, LiDAR point clouds, and manually annotated ground-truth data. Data for the dataset was collected from arid cotton fields across Australia via UAV and LiDAR technologies during its construction. Primarily applied to crop biomass prediction, the MMCBE dataset is also utilized to explore computer vision tasks such as 3D crop reconstruction and novel view synthesis, aiming to address the challenges of high-resolution crop biomass estimation.




