情報処理学会 第79回全国大会講演要旨

In the recent years, LiDAR devices are being widely used in reconstruction of the environment. Especially, mobile 3D LiDAR sensors like Velodyne HDL-32E are attracting more interests for these mobility, portability and the low cost compared to the stationary LiDAR sensors. In this work, we discuss an approach for automatically reconstructing the indoor environment both on point cloud level and surface level with the raw data acquired by a Velodyne HDL-32E LiDAR. Our approach consists of many existing methods of registration, segmentation, surface fitting and surface reconstruction. The approach is implemented and applied on the real-word of the indoor environment, which is acquired with the Velodyne HDL-32E LiDAR in the exhibition area of Geo spatial EXPO 2016 in Tokyo. The processed result shows the feasibility and the practicability of the proposed application.