点云训练样本制作中LAS与HDF5数据格式转换

Data format conversion of LAS and HDF5 in the production of point cloud training samples

深度学习技术在三维激光点云数据处理中的应用是近年来新的研究热点,应用点云数据(LAS格式)制作深度学习训练样本(HDF5文件格式)过程中涉及两种文件格式的转换的问题。基于VS2013开发平台创建MFC工程,搭建LAStools、HDF_Group工程环境,利用车载移动测量系统采集到的LAS格式数据,设计了一款数据格式转换程序。经过实验验证,该转换程序稳定有效,解决了点云训练样本制作过程中数据格式转换的问题。

Application of deep learning technology in 3D laser point cloud data processing was a new research hotspot in recent years. There was a problem involved in the conversion of two file formats in the process of creating deep learning training samples （HDF5 file format） using point cloud data （LAS format） . A data format conversion program was designed using LAS format data collected by mobile measurement system based on the VS2013 development platform to create an MFC project. Experiment verified that the conversion program was stable and effective solving the problem of data format conversion in the production of point cloud training samples.