“3D点云处理”
参考资料：
- 3D点云综述
- 深度学习在3D点云处理中的探索
- 3D is here: Point Cloud Library (PCL)
点云是什么
- 某个坐标系下点的数据集
- 包含了丰富的信息，包括三维坐标X，Y，Z、颜色、分类值、强度值、时间等
获取途径
主要是通过三维激光扫描仪进行数据采集获取点云数据，其次通过二维影像进行三维重建，在重建过程中获取点云数据，另外还有一些，通过三维模型来计算获取点云。
Light Detection And Ranging (LIDAR)
Intro
- 点云深度学习包括：3D形状分类，3D对象检测和跟踪以及3D点云分割(3D shape classification, 3D object detection and tracking, and 3D point cloud segmentation)
- 3D数据的存储格式：深度图像，点云，网格和体积网格；点云表示将原始几何信息保留在3D空间中，而不会进行任何离散化
- 在3D点云上进行深度学习仍然面临数个重大挑战，例如数据集规模小，维数高和3D点云的非结构化性质
PCL
Point Cloud Library
PCL provides support for all the common 3D building blocks that applications need.
The library contains state-of- the art algorithms for: filtering, feature estimation, surface reconstruction, registration, model fitting and segmentation. PCL is supported by an international community of robotics and perception researchers(点云获取、滤波、分割、配准、检索、特征提取、识别、追踪、曲面重建、可视化).
- Robots need to be able to perceive the world
- PCL helps robots and provides mechanism for handling point clouds efficiently
- PCL is for n-D Point Clouds and 3D geometry processin
- PCL is fully integrated with ROS, the Robot Operating System (see http://ros.org)
