WEKO3
アイテム
車載カメラを用いた障害物検出システムの研究
https://doi.org/10.18997/00004096
https://doi.org/10.18997/000040968d41abf0-0c30-4dfc-ab67-dd374acffed3
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
D-231_kou_k_371.pdf (1.2 MB)
|
|
| アイテムタイプ | 学位論文 = Thesis or Dissertation(1) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 公開日 | 2014-12-03 | |||||||||
| 資源タイプ | ||||||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_db06 | |||||||||
| 資源タイプ | doctoral thesis | |||||||||
| タイトル | ||||||||||
| タイトル | A Study on an Obstacles Detection System Employing a Car-mounted Camera | |||||||||
| 言語 | en | |||||||||
| タイトル | ||||||||||
| タイトル | 車載カメラを用いた障害物検出システムの研究 | |||||||||
| 言語 | ja | |||||||||
| 言語 | ||||||||||
| 言語 | eng | |||||||||
| 著者 |
銭, 少華
× 銭, 少華
|
|||||||||
| 抄録 | ||||||||||
| 内容記述タイプ | Abstract | |||||||||
| 内容記述 | In recent years, autonomous collision avoidance systems have been researched and developed for realizing safe driving using cameras and sensors. These systems are designed to warn the drivers the presence of obstacles on the road and help them take a necessary action in advance. In these systems, the ability to detect obstacles is essential. Although various methods of obstacles detection have already been reported, these existing obstacles detection methods have some inadequacies: Some of them can be only used to detect moving obstacles; Some of them cannot extract the shape of obstacles, and they only use a rectangular frame that surrounds an obstacle to represent a detected obstacle; Some of them can only be used to detect one kind of specific object, such as pedestrian detection or vehicle detection.In order to make up for the inadequacies of the existing obstacles detection method, in this thesis, a method is proposed for detecting obstacles on a road by the employment of the background modeling and the road region detection.In obstacles detection, true obstacles are defined as arbitrary objects which protrude from the ground plane in the road region, including static and moving objects. Road marks in the road region and objects outside the road region are considered as false obstacles. The output of this obstacles detection method is based on the obstacles’ shape.In this thesis, we also propose a method of classifying 2D objects and 3D objects. The results of 2D objects and 3D objects classification can be used in the resultant image of obstacles detection to delete 2D objects （such as road marks） and improve the accuracy of obstacles detection.The originalities of this thesis are as follows:In the first place, the proposed method can detect arbitrary objects including both static objects and moving objects. This is helpful because static objects such as boxes fallen on the road from a car are dangerous for drivers.In the second place, the output of the proposed method is the shape of obstacles. Extraction of the shape of an obstacle is important for obstacles recognition. If the detected obstacle is recognized as a pedestrian from its shape, we can foresee his/her next motion. In the third place, the proposed method can distinguish which objects are 3D objects, and which objects are 2D objects in a pile of objects using a monocular camera. It is useful in the obstacles detection and other applications, such as navigation of walking robots.In the performed experiments, it is shown that the proposed obstacles detection method is able to extract the shape of both static and moving obstacles in a frontal view from a car. | |||||||||
| 目次 | ||||||||||
| 内容記述タイプ | TableOfContents | |||||||||
| 内容記述 | 1 Introduction||2 Obstacles Detection||3 2D and 3D Objects Classification||4 Final Experimental Results and Evaluation||5 Conclusion||References | |||||||||
| 備考 | ||||||||||
| 内容記述タイプ | Other | |||||||||
| 内容記述 | 九州工業大学博士学位論文 学位記番号:工博甲第371号 学位授与年月日:平成26年9月26日 | |||||||||
| キーワード | ||||||||||
| 主題Scheme | Other | |||||||||
| 主題 | 障害物検出 | |||||||||
| キーワード | ||||||||||
| 主題Scheme | Other | |||||||||
| 主題 | 移動カメラ | |||||||||
| キーワード | ||||||||||
| 主題Scheme | Other | |||||||||
| 主題 | 背景推定 | |||||||||
| キーワード | ||||||||||
| 主題Scheme | Other | |||||||||
| 主題 | 道路検出 | |||||||||
| キーワード | ||||||||||
| 主題Scheme | Other | |||||||||
| 主題 | GMM | |||||||||
| アドバイザー | ||||||||||
| タン, ジュークイ | ||||||||||
| 学位授与番号 | ||||||||||
| 学位授与番号 | 甲第371号 | |||||||||
| 学位名 | ||||||||||
| 学位名 | 博士（工学） | |||||||||
| 学位授与年月日 | ||||||||||
| 学位授与年月日 | 2014-09-26 | |||||||||
| 学位授与機関 | ||||||||||
| 学位授与機関識別子Scheme | kakenhi | |||||||||
| 学位授与機関識別子 | 17104 | |||||||||
| 学位授与機関名 | 九州工業大学 | |||||||||
| 学位授与年度 | ||||||||||
| 内容記述タイプ | Other | |||||||||
| 内容記述 | 平成26年度 | |||||||||
| 出版タイプ | ||||||||||
| 出版タイプ | VoR | |||||||||
| 出版タイプResource | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |||||||||
| アクセス権 | ||||||||||
| アクセス権 | open access | |||||||||
| アクセス権URI | http://purl.org/coar/access_right/c_abf2 | |||||||||
| ID登録 | ||||||||||
| ID登録 | 10.18997/00004096 | |||||||||
| ID登録タイプ | JaLC | |||||||||
