2013 Fiscal Year Final Research Report
Automatic Mesh Modeling of Large-Scale Environments based on Accurate Recognition of Laser-Scanned Point Clouds
| Project/Area Number |
24760114
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Design engineering/Machine functional elements/Tribology
|
|---|
| Research Institution | Hokkaido University |
|---|
Principal Investigator |
DATE Hiroaki 北海道大学, 情報科学研究科, 准教授 (20374605)
|
|---|
| Project Period (FY) |
2012-04-01 – 2014-03-31
|
| Keywords | レーザ計測 / 点群 / メッシュモデル / レジストレーション / セグメンテーション / 物体認識 |
|---|
| Research Abstract |
In this research, a set of point cloud data processing algorithms for generating mesh models of large-scale environments from point clouds acquired by laser scanning was developed. In particular, the following algorithms were developed: (1) efficient registration algorithms of point clouds of large-scale environments, (2) an effective point cloud rendering algorithm to support easy visual understanding of the scanned environments, (3) a segmentation algorithm of point clouds of urban environments, (4) pole-like object recognition and classification algorithms from point clouds, and (5) mesh modeling algorithms of the objects in urban environments. The effectiveness of each algorithm was evaluated through the experiments using several point clouds, and it was confirmed that the mesh models of the pole-like objects and street trees could be generated efficiently from point clouds of urban environments using the developed algorithms.
|
