源码笔记记录
1.Boundary-refined supervoxels 边界细化超体素
链接:https://github.com/RioWong/BRSS
作者:RioWong
Boundary-refined supervoxels for point cloud segmentation It is three-step method:
- get boundary-refined supervoxels based on the improved VCCS
- get facets based on region growing with improved seed selection and growth criteria
- get object-level segmentation by convexly merging
用于点云分割的边界细化超体素是三步法:
- 基于改进的 VCCS 获得边界细化的超体素
- 通过改进的种子选择和生长标准获得基于区域生长的方面
- 通过凸合并获得对象级分割
2.Toward better boundary preserved supervoxel segmentation for 3D point clouds.
链接:https://github.com/yblin/Supervoxel-for-3D-point-clouds
作者:yblin
引用:Lin Y, Wang C, Zhai D, W Li, and J Li. Toward better boundary preserved supervoxel segmentation for 3D point clouds. Isprs Journal of Photogrammetry & Remote Sensing, vol. 143, pages 39-47, 2018.
注意:QT环境运行
3.Global L0 algorithm for regularity-constrained plane fitting
链接:https://github.com/yblin/global_l0
作者:yblin
引用:Lin Y, Li J, Wang C, Chen Z, Wang Z, and Li J. Fast regularity-constrained plane fitting. Isprs Journal of Photogrammetry & Remote Sensing, vol. 161, pages 208-217, 2020.
我们引入了一种基于能量最小化的方法来重建与我们的约束模型一致的平面。所提出的算法高效、易于理解且易于实现。实验结果表明,我们的算法成功地拟合了高百分比噪声和异常值下的平面。这在速度和鲁棒性方面优于其他最先进的规律性约束平面拟合方法。
注意:平面提取
4.Fast 3D point cloud segmentation using supervoxels with geometry and color for 3D scene understanding
链接:https://github.com/fverdoja/Fast-3D-Pointcloud-Segmentation
作者:fverdoja
引用:F. Verdoja, D. Thomas, and A. Sugimoto, “Fast 3D point cloud segmentation using supervoxels with geometry and color for 3D scene understanding,” in IEEE International Conference on Multimedia and Expo (ICME 2017), Hong Kong, 2017, pp. 1285–1290.
3D 彩色点云的分割是一个重新引起人们兴趣的研究领域,这要归功于最近廉价的消费 RGB-D 相机的可用性及其在许多机器人应用中作为不可避免的低级步骤的重要性。然而,3D 数据的性质使任务具有挑战性,因此,许多不同的技术被提出,所有这些都需要昂贵的计算成本。本文提出了一种新的快速 3D 彩色点云分割方法。它从云的超体素分割开始,即对云中的点进行过分割。然后,它利用一种利用几何和颜色的新度量来迭代合并超体素,以获得保持分区层次结构的 3D 分割。该算法还呈现与输入大小成线性关系的计算复杂度。两个公开可用数据集的实验结果表明,我们提出的方法优于最先进的技术。
命令行:./supervoxel_clustering -p <path/to/milk.pcd> -t 0.2 --AL
5.Boundary Preserving SuperVoxel Clustering with CUDA acceleration
链接:https://github.com/shishenghuang/SuperVoxelCuda
作者:Shi-Sheng Huang
使用 CUDA 加速的边界保持超体素聚类
6.SalientSupervoxel
链接:https://github.com/GeeeG/SalientSupervoxel#salientsupervoxel
作者:GeeeG
引用:Further details are available in our paper on the subject. If you use this code in an academic context, please cite the paper: Ge Gao, Mikko Lauri, Jianwei Zhang and Simone Frintrop. “Saliency-guided adaptive seeding for supervoxel segmentation”, IROS 2017.
这是用于超体素分割 (SSV) 的显着性引导自适应播种的 C++ 实现。
基本思想是通过使用视觉显着性作为先验知识,在更显着的区域中生长更小的超体素,在不太显着的区域中生长更大的超体素。
左侧是输入点云,中间是使用 VCCS 进行统一播种的过分割结果,右侧是使用我们的方法 (SSV) 进行显着性引导自适应播种的过分割结果。
7.CSF平面分割
链接:https://github.com/jianboqi/CSF
Airborne LiDAR filtering method based on Cloth Simulation. This is the code for the article:
W. Zhang, J. Qi*, P. Wan, H. Wang, D. Xie, X. Wang, and G. Yan, “An Easy-to-Use Airborne LiDAR Data Filtering Method Based on Cloth Simulation,” Remote Sens., vol. 8, no. 6, p. 501, 2016. (http://www.mdpi.com/2072-4292/8/6/501/htm)
8.Real-Time Detection of Planar Regions in Unorganized Point Clouds平面分割
链接:https://www.inf.ufrgs.br/~oliveira/pubs_files/HT3D/HT3D_page.html
Automatic detection of planar regions in point clouds is an important step for many graphics, image processing, and computer vision applications. While laser scanners and digital photography have allowed us to capture increasingly larger datasets, previous techniques are computationally expensive, being unable to achieve real-time performance for datasets containing tens of thousands of points, even when detection is performed in a non-deterministic way. We present a deterministic technique for plane detection in unorganized point clouds whose cost is O ( n log n ) in the number of input samples. It is based on an efficient Hough-transform voting scheme and works by clustering approximately co-planar points and by casting votes for these clusters on a spherical accumulator using a trivariate Gaussian kernel. A comparison with competing techniques shows that our approach is considerably faster and scales significantly better than previous ones, being the first practical solution for deterministic plane detection in large unorganized point clouds.
9. Segment cuboid-shaped objects (e.g. parcels) without any task specific training data
链接:https://github.com/a-nau/Plane-Segmentation-Refinement
10. 语义分割标注工具
链接:https://blog.csdn.net/qq_27353621/article/details/122049401?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522164915599816781685363404%2522%252C%2522scm%2522%253A%252220140713.130102334.pc%255Fall.%2522%257D&request_id=164915599816781685363404&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2allfirst_rank_ecpm_v1~rank_v31_ecpm-4-122049401.142v5control,157v4control&utm_term=txt%E8%BD%ACXYZL&spm=1018.2226.3001.4187
搜索关键词
supervoxel
plane detedtion
indoor point cloud segmentation
dbscan