点云配准资源汇总

点云配准的目标是根据原始点云和目标点云，通过配准求出变换矩阵，即旋转矩阵R和平移矩阵T，并计算误差，来比较匹配结果。主要有以下几种比较

• 基于局部特征描述子（PFH、FPFH、3Dsc，Shot等等）；
• icp配准 ；
• 基于概率分布 （NDT）；

配准的一般步骤：

• 提取关键点
• 特征描述
• 一致性估计（以上可以概括为粗配准）
• 精配准
• 误差分析

注意：配准中，由于不同点云数据集的特性，需要提取不同关键点。

demo展示

粗配准

ICP迭代配准

汇总ICP资源

1，FilterReg: Robust and Efficient Probabilistic Point-Set Registration using Gaussian Filter and Twist pdf:Parameterization pdf:https://arxiv.org/pdf/1811.10136.pdf

2，Robust Point Cloud Registration Using Iterative Probabilistic Data Associations ("Robust ICP") code: https://github.com/ethz-asl/robust_point_cloud_reg

3，Efficient Global Point-cloud registration code:https://github.com/nmellado/Super4PCS

4，Scale Ratio ICP for 3D Point Clouds with Different Scales

code:https://github.com/linbaowei/ScaleRatioICP

5，improved ICP for partial overlapping area situation code: https://github.com/jieliu/point_cloud_registration

6，使用ICP实现点云的配准与拼接的SLAM算法

code：https://github.com/kzampog/kabamaru

7,Fast Point Feature Histograms (FPFH) for 3D Registration.

PCL点云配准（1）

PCL点云配准（2）

8,3D Point Cloud Registration for Localization using a Deep Neural Network Auto-Encoder.

code：https://github.com/gilbaz/LORAX

9，Density Adaptive Point Set Registration.

code：https://github.com/felja633/DARE

10，Fast rotation search with stereographic projections for 3D registration

code：https://cs.adelaide.edu.au/~aparra/project/pcr/

11，Discriminative Optimization: Theory and Applications to Point Cloud Registration pdf:https://www.researchgate.net/publication/320964941_Discriminative_Optimization_Theory_and_Applications_to_Point_Cloud_Registration

12，Markerless point cloud registration with keypoint-based 4-points congruent sets

pdf:https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/II-5-W2/283/2013/

13,Comparing ICP Variants on Real-World Data Sets code:https://github.com/ethz-asl/libpointmatcher

……