(18.207.253.100) 您好!臺灣時間:2021/05/06 08:40
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:葉雲豪
研究生(外文):Yun-Hao, Yeh
論文名稱:三維物件之點雲模型編輯與辨識系統
論文名稱(外文):3D Object Point Cloud Model Editing and Recognition System
指導教授:張厥煒張厥煒引用關係
指導教授(外文):Chueh-Wei, Chang
口試委員:奚正寧楊士萱
口試日期:2016-07-15
學位類別:碩士
校院名稱:國立臺北科技大學
系所名稱:資訊工程系研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
畢業學年度:104
語文別:中文
中文關鍵詞:三維物件辨識特徵抽取點雲函式庫
外文關鍵詞:3D Object RecognitionFeature ExtractPoint Cloud Library (PCL)
相關次數:
  • 被引用被引用:0
  • 點閱點閱:416
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:25
  • 收藏至我的研究室書目清單書目收藏:1
本系統第一目標為從一開始的三維影像輸入後進行編輯到最後的三維物件辨識,可以由本系統全部完成,第二目標為將各階段分開,讓開發者可以根據不同需求,針對各部分繼續進行開發,且可以讓開發者在不同作業平台上快速上手繼續開發;此外也可以有不同的搭配組合,找出最適合三維物件辨識的組合。
本系統功能分為三部分,第一部分為點雲模型的編輯,利用PCL(Point Cloud Library, PCL)提供的各模組將輸入的點雲資料進行過濾雜訊像是Region Filter、 Statistical Outlier Removal、 Radius Outlier Removal、 Median Filter,內容分割像是Color Segmentation、Parametric Model Segmentation,疊合像是ICP Registration,最後能建立一個點雲模型;第二部份為特徵訓練,負責抽取關鍵點與抽取特徵,並以像是Normal或是SHOT等特徵描述方式儲存進點雲模型特徵資料庫,第三部分為特徵比對,負責從點雲模型特徵資料庫裡讀取特徵,利用Correspondence Grouping進行三維物件辨識。
最後本系統完成了編輯點雲模型,儲存進點雲模型特徵資料庫,建立了點雲模型特徵資料庫,並且可以從資料庫讀取特徵進行三維物件辨識,完成本系統的目標。在三維物件辨識上,利用SHOT特徵描述方式搭配Correspondence Grouping,在旋轉角度變化不大的情況下,可以有好的辨識率。
The first goal of our system is that use this system to edit the three dimension image which loaded from the sensor. Then the three dimension point cloud model can be recognized. The second goal of our system is that separate the system into three independent parts. The developer can keep developing system depend on what they need. For each part, the developer can choice the more suitable combination of three dimension object recognition method.
This system has three parts. First part is the edit of point cloud model. In first part our system use the PCL(Point Cloud Library, PCL) module to provide the noise filter such as Region Filter, Statistical Outlier Removal, Radius Outlier Removal, and Median Filter, the model segment such as Color Segmentation, and Parametric Model Segmentation, registration such as ICP Registration, which construct a point cloud model. Second part is feature training, which is responsible for extracting keypoints and feature, with feature descript such as Normal, SHOT and store the data in the three-dimensional point cloud model feature database. Third part is feature recognize, which is responsible for loading feature from the three dimensional point cloud model feature database, than use correspondence grouping to recognize the three dimensional object.
Our system support the edit point cloud model, and save the data into feature database, build the point cloud model feature database, them use the feature from the point cloud model feature database to recognize the three dimension object. We use SHOT feature descript and correspondence grouping, can get a good recognition rate in the situation that the rotation angle of object is small.
目 錄

摘 要 i
ABSTRACT ii
誌 謝 iv
目 錄 v
表目錄 viii
圖目錄 ix
第一章 緒論 1
1.1 研究動機 1
1.2 研究目的 1
1.3 論文架構 2
第二章 相關研究與文獻探討 3
2.1 點雲相關文獻 3
2.1.1 點雲模型介紹 3
2.1.2 點雲函式庫介紹 3
2.2 三維點雲模型編輯相關文獻 8
2.3 三維物件辨識相關文獻 11
2.4 總結 11
第三章 系統架構與流程 12
3.1 系統架構 12
3.2 點雲模型編輯模組 13
3.3 特徵訓練模組 13
3.4 特徵比對模組 14
3.5 設計理念 15
第四章 點雲模型編輯模組 18
4.1 檔案子系統(File Subsystem) 18
4.1.1 輸入/輸出功能 18
4.1.2顯示功能 19
4.1.3反轉(Inverse)功能 19
4.2 過濾子系統(Filter Subsystem) 21
4.2.1 Region Filter 21
4.2.2 Statistical Outlier Removal 21
4.2.3 Radius Outlier Removal 23
4.2.4 Median Filter 23
4.2.5 過濾子系統總結 24
4.3 分割子系統(Segmentation Subsystem) 25
4.3.1 顏色分割(Color Segmentation) 25
4.3.2參數模型分割(Parametric Model Segmentation) 26
4.3.3 分割子系統總結 27
4.4 疊合子系統(Registration Subsystem) 28
4.4.1 疊合子系統總結 30
第五章 特徵訓練模組與特徵比對模組 31
5.1 特徵訓練模組 31
5.1.1關鍵點 31
5.1.2特徵 32
5.1.3特徵描述方式 33
5.2 特徵比對模組 34
5.2.1 Correspondence 35
5.2.2 Grouping 35
第六章 實驗結果 36
6.1 點雲編輯模組介面 36
6.1.1檔案子系統(File Subsystem) 36
6.1.2過濾子系統(Filter Subsystem) 37
6.1.3分割子系統(Segmentation Subsystem) 38
6.1.4疊合子系統(Registration Subsystem) 39
6.2 特徵訓練模組與特徵比對模組介面 40
6.3 測試點雲模型編輯模組 41
6.4 測試特徵抽取模組與特徵比對模組 52
6.5 總結 56
第七章 結論與未來展望 57
7.1 結論 57
7.2 未來展望 58
參考文獻 59
參考文獻

[1]PCL - Point Cloud Library (PCL), http://pointclouds.org/about/.
[2]Point Cloud Library - Walk Through, http://pointclouds.org/documentation/tutorials/walkthrough.php#walkthrough.
[3]Point Cloud Library – In Hand Scanner http://pointclouds.org/documentation/tutorials/in_hand_scanner.php.
[4]3D System Sense, http://www.3dsystems.com/shop/realsense/sense.
[5]3D Point Cloud Editor, http://paradise.caltech.edu/~yli/software/pceditor.html.
[6]MeshLab, http://meshlab.sourceforge.net/.
[7]A. Hoover, G. Jean-Baptiste, X. Jiang, P. J. Flynn, H. Bunke, D. B. Goldgof, K. Bowyer, D. W. Eggert, A. Fitzgibbon, and R. B. Fisher, “A experimental comparison of range image segmentation algorithms,” IEEE Tran. Pattern Analysis and Machine Intelligence, vol. 18, no. 7, pp. 673-689, July 1996.
[8]G. Medioni and B. Parvin, “Segmentation of range images into planar surfaces by split and merge,” IEEE Conf. Computer Vision and Pattern Recognition, pp. 415–417, 1986.
[9]R. L. Hoffman and A. K. Jain, “Segmentation and classification of range images,” IEEE Tran. Pattern Analysis and Machine Intelligence, vol. 9, no. 5, pp. 608-620, Sept. 1987.
[10]P. J. Flynn and A. K. Jain, “BONSAI: 3D object recognition using constrained search,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 13, no. 10.
[11]J. M. Jolion, P. Meer, and S. Bataouche, “Robust clustering with applications in computer vision,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 13, no. 8, Aug. 1991.
[12]A. Bab-Hadiashar and N. Gheissari, “Range image segmentation using surface selection criterion,” IEEE Trans. on Image Processing, vol. 15, no. 7, July 2006.
[13]T. U. Fan, G. Medioni, and R. Nevatia, “Recognizing 3-D objects using surface descriptions,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 11, pp. 1140-1157, Nov.1989.
[14]Lihui Wang, Baozong Yuan, “Curvature and density based feature point detection for point cloud data,” IET 3rd International Conference. Wireless, Mobile and Multimedia Networks (ICWMNN 2010), pp26-29, Sept 2010.
[15]Aitor Aldoma, Zoltan-Csaba Marton, Federico Tombari, Walter Wohlkinger, Christian Potthast, Bernhard Zeisl, Radu Bogdan Rusu, Suat Gedikli, Markus Vincze, “Tutorial: point cloud library: three-dimensional object recognition and 6 DOF Pose Estimation,” IEEE Robotics & Automation Magazine, Volume:19, Issue: 3, Sept. 2012.
[16]QT, https://www.qt.io/
[17]OpenNI 的座標系統| Heresy’s Space, https://kheresy.wordpress.com/2012/04/05/coordinate-system-in-openni/.
[18]Point Cloud Library - Statistical Outlier Removal http://pointclouds.org/documentation/tutorials/statistical_outlier.php#statistical-outlier-removal.
[19]Point Cloud Library - Radius Outlier Removal http://pointclouds.org/documentation/tutorials/remove_outliers.php#remove-outliers.
[20]Ruwen Schnabel, Roland Wahl, Reinhard Klein, “Efficient RANSAC for point cloud shape detection,” Universitat Bonn, Computer Graphics Group, vol.0, no0, 1981, pp. 1-12.
[21]P. J. Besl and N. D. McKay, “A method for registration of 3-D shapes,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.14, no2, 1992, pp. 239-256.
[22]Point Cloud Library –Interactive ICP http://pointclouds.org/documentation/tutorials/interactive_icp.php#interactive-icp.
[23]Point Cloud Library – Normal Feature http://pointclouds.org/documentation/tutorials/normal_estimation.php#normal-estimation.
[24]Point Cloud Library – OpenNI tutorial 4: 3D object recognition (descriptors) http://robotica.unileon.es/mediawiki/index.php/PCL/OpenNI_tutorial_4:_3D_object_recognition_(descriptors)#SHOT.
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關論文
 
無相關期刊
 
無相關點閱論文
 
系統版面圖檔 系統版面圖檔