| Project/Area Number |
17K00363
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent robotics
|
|---|
| Research Institution | Kobe University |
|---|
Principal Investigator |
Tazaki Yuichi 神戸大学, 工学研究科, 准教授 (10547433)
|
|---|
| Project Period (FY) |
2017-04-01 – 2020-03-31
|
| Project Status |
Completed (Fiscal Year 2019)
|
| Budget Amount *help |
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2019: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2018: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
Fiscal Year 2017: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
|
|---|
| Keywords | 移動ロボット / 地図生成 / 自己位置推定 / SLAM / LiDAR / ループ検出 / 実時間情報処理 / デジタル地図 |
|---|
| Outline of Final Research Achievements |
This research focused on digital map, which is a crucial element for realizing intelligent mobile robots and autonomously driving cars, and aimed at revealing the mathematical representation and characteristics of maps that are not only precise enough to capture the real world but also easy to use for the robots from the viewpoint of data size and efficiency of computation. More concretely, this research proposed a special keypoint called proximity point that can be extracted from pointcloud data produced by 3D laser range finders (3D LiDARs). By detecting proximity points from pointcloud data, the data size can be reduced dramatically while preserving the characteristics of the observed scene. Based on the geometric characteristics of proximity points, a number of techniques
necessary for realizing autonomous navigation including loop detection, self-localization, and moving object detection were developed and their performance was tested in field experiments.
|
| Academic Significance and Societal Importance of the Research Achievements |
本研究の学術的な意義は，三次元点群データから単純なアルゴリズムにより抽出可能な特徴点を提案し，その特徴点が持つ幾何学的性質を明らかにした点にある．また，その性質を基礎として観測データ間の類似性を特徴点に基づいて評価する方法を提案し，ループ検出や自己位置推定などの種々の手法へ展開したことに意義が認められる．社会的な意義としては，近年急速な普及を見せているものの処理コストの高さが課題であった3DLiDARに対して，膨大な点群データを比較的小さな計算コストで処理する方法論を明らかにしたことで，ナビゲーションシステムの組み込み計算機などへの実装が可能となり，移動ロボットの低コスト化に寄与した点にある．
|
