树结构点云数据集
收藏资源简介:
树结构点云数据集由重庆大学计算机科学与技术学院创建,包含900个树模型及其对应的地面实况骨架。数据集涵盖了不同规模、分支密度和复杂度的树模型,模拟了自然环境中的点云扫描,挑战了典型的骨架提取算法。创建过程中,首先使用树编辑器构建树骨架和相应的网格模型,然后使用隐式表面模拟距离扫描以获取点云。数据集可用于评估骨架提取算法的性能,通过比较提取的骨架与地面实况骨架来进行算法间的评估。
This tree-structured point cloud dataset was created by the School of Computer Science and Technology, Chongqing University. It contains 900 tree models and their corresponding ground-truth skeletons. The dataset covers tree models with varying scales, branch densities and complexities, simulating point cloud scans in natural environments and posing challenges to typical skeleton extraction algorithms. During its creation, tree skeletons and their corresponding mesh models were first built using a tree editor, then point clouds were obtained by simulating distance scans via implicit surfaces. This dataset can be used to evaluate the performance of skeleton extraction algorithms, enabling inter-algorithm comparison by contrasting extracted skeletons with the ground-truth ones.




