BioVista
收藏资源简介:
BioVista数据集由丹麦奥尔堡大学等机构创建,旨在通过多模态数据融合提升森林生物多样性评估的准确性。该数据集包含44,378对2D正射影像和3D激光雷达点云数据,覆盖丹麦温带森林区域。正射影像分辨率为12.5厘米,激光雷达点云密度为8点/平方米。数据集创建过程包括从高自然价值(HNV)森林地图中筛选样本,并确保正射影像与激光雷达数据的时间一致性。该数据集的应用领域主要集中在森林生态系统管理,旨在通过深度学习模型结合光谱和结构信息,解决大规模生物多样性评估的难题。
The BioVista dataset was created by Aalborg University, Denmark, and other institutions, aiming to improve the accuracy of forest biodiversity assessment through multimodal data fusion. It contains 44,378 pairs of 2D orthophotos and 3D LiDAR point cloud data, covering temperate forest regions in Denmark. The orthophotos have a resolution of 12.5 cm, while the LiDAR point clouds have a density of 8 points per square meter. The dataset creation process involves screening samples from High Nature Value (HNV) forest maps and ensuring temporal consistency between the orthophotos and LiDAR data. The primary application domains of this dataset focus on forest ecosystem management, and it aims to address the challenges of large-scale biodiversity assessment by combining spectral and structural information with deep learning models.




