多深度攝影機之視野對齊｜國立臺灣科技大學博碩士論文系統

簡易檢索 / 詳目顯示

回結果列表

研究生：	胡維中 Wei-Chung Hu
論文名稱：	多深度攝影機之視野對齊 Field of View Alignment with Multiple RGB-D Cameras
指導教授：	林昌鴻 Chang-Hong Lin
口試委員:	阮聖彰 Shanq-Jang Ruan　李佳翰 Chia-han Lee 沈毅偉 Yi-Wei Shen
學位類別：	碩士 Master
系所名稱：	電資學院 - 電子工程系 Department of Electronic and Computer Engineering
論文出版年：	2013
畢業學年度：	101
語文別：	英文
論文頁數：	64
中文關鍵詞：	視野、對齊、深度攝影機、特徵匹配、三維註冊、座標系統轉換
外文關鍵詞：	Field of View, Alignment, RGB-D Camera, Feature Matching, 3D Registration, Coordinate System Transformation
相關次數：	點閱：566 下載：9
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

上一筆

自從深度攝影機被發表之後，例如：Kinect與Xtion，有越來越多的體感應用被發展。然而，深度攝影機的水平視野只有57.5°，當多人一起操控深度攝影機的應用時，或是在一較大的環境時，深度攝影機的使用會受到限制。因此本論文使用多深度攝影機來產生一較廣的視野。我們的方法分成兩個步驟，轉換矩陣生成以及深度資料對齊。轉換矩陣生成是一個初始化的步驟，在這個步驟中，我們的系統會在深度攝影機重疊的視野中找出移動物體，並從移動物體中擷取出匹配的特徵點，並利用匹配的特徵點計算出深度資料在世界座標系統中的轉換矩陣。轉換矩陣產生後，系統將會對齊接下來世界座標系統中的深度資料，並將對齊好的資料轉換為標準的深度資料格式。在實驗結果中能發現我們的系統在多種情況下有效的增加視野。較廣的視野能用來追蹤或辨識人體骨架給其他體感應用來使用。

Since RGB-D cameras, such as Kinect and Xtion, have been released, many motion based applications have been developed. However, the horizontal field of view (FOV) of an RGB-D camera is only 57.5°, which limits the usage when there are many people interact the applications or in a large scale environment. In this thesis, we proposed a method to generate a wider FOV with multiple RGB-D cameras. There are two main steps in the proposed method, including transformation matrix generation and depth data alignment. The transformation matrix generation serves as an initial step. In this step, our system would extract the matching feature points from the moving object in the overlapped region. The matching feature points can be used to calculate the transformation matrix between the data of two cameras in the world coordinate system. After initialization, we use the generated transformation matrix to align the data in the world coordinate system from both cameras for all the following frames. The aligned data would be formed a standard depth data format. The results shows that our system can increase the effective FOV in the different conditions. The wider FOV can then be used to recognize the skeleton gestures or track the skeleton for other applications.

摘要................................................................................................................................ I
Abstract ......................................................................................................................... II
致謝.............................................................................................................................. III
List of Contents ............................................................................................................ IV
List of Figures .............................................................................................................. VI
List of Tables ................................................................................................................ IX
CHAPTER 1 INTRODUCTION ................................................................................... 1
1 Motivation ........................................................................................................ 1
2 Contribution ..................................................................................................... 2
3 Thesis Organization ......................................................................................... 2
CHAPTER 2 RELATED WORKS ................................................................................ 3
1 Correspondence................................................................................................ 3
1.1 Correlation-based Method .................................................................... 3
1.2 Feature-based Method ........................................................................... 4
2 3D Registration ................................................................................................ 5
2.1 3D Laser Scanning Data ....................................................................... 5
2.2 Light Coding Data................................................................................. 6
CHAPTER 3 PROPOSED METHODS ........................................................................ 9
1 Coordinate System Transformation of RGB-D Camera .................................. 9
2 Framework ..................................................................................................... 14
3 Transformation Matrix Generation ................................................................ 15
3.1 Preprocessing ...................................................................................... 16
3.2 Moving Object Finding ....................................................................... 17
3.3 Feature Matching ................................................................................ 21
3.4 Transformation Matrix Generation ..................................................... 25
4 Depth Data Alignment ................................................................................... 31
4.1 Projective to World Transformation .................................................... 31
4.2 Alignment ............................................................................................ 32
4.3 Perspective Projection and Hidden Surface Removal ........................ 36
CHAPTER 4 EXPERIMENTAL RESULTS ............................................................... 39
1 Developing Platform ...................................................................................... 39
2 Experimental Results ..................................................................................... 40
2.1 Experiment 1 ....................................................................................... 40
2.2 Experiment 2 ....................................................................................... 44
2.3 Experiment 3 ....................................................................................... 45
V
2.4 Experiment 4 ....................................................................................... 47
3 Analysis of Proposed System ......................................................................... 50
3.1 Analysis of Alignment......................................................................... 50
3.2 Performance Evaluation of Transformation Matrix Generation ......... 53
3.3 Performance Evaluation of Depth Data Alignment ............................ 57
CHAPTER 5 CONCLUSIONS AND FUTURE WORKS .......................................... 59
1 Conclusions .................................................................................................... 59
2 Future Works .................................................................................................. 60
References .................................................................................................................... 61

                                

[1] Microsoft. (2013). Kinect [Online]. http://msdn.microsoft.com/en-us/library/hh85-
5347.aspx

[2] ASUSTeK Computer Inc. (2013). Xtion PRO LIVE [Online]. http://www.asus.co-
m/Multimedia/Xtion_PRO_LIVE

[3] UNITED ARROWS. (2013). Marionettebot [Online]. http://koisuru.jp/m/

[4] LANYANG MUSEUM. (2013). "Birth of Yilan" interactive theatre [Online]. http://www.lym.gov.tw/eng/en_infor_2.asp

[5] 台北榮民總醫院. (2013). 虛擬實境復健治療 [Online]. http://homepage.vghtp- e.gov.tw/~nrl/service4.htm

[6] Microsoft. (2011). Kinect for Windows 創意應用競賽決賽得獎作品 [Online]. http://www.youtube.com/watch?v=1u4kFfhCgyY

[7] K. Nandakumar; A. K. Jain, "Local Correlation-based Fingerprint Matching," in Indian Conference on Computer Vision, Graphics and Image Processing, 2004, pp. 503-508.

[8] Y. Tsin; T. Kanade, "A Correlation-Based Approach to Robust Point Set Registration," in Computer Vision-ECCV 2004, 2004, pp. 558-569.

[9] H. Hirschmuller, "Improvements in Real-Time Correlation-Based Stereo Vision," in Stereo and Multi-Baseline Vision, 2001.(SMBV 2001). Proceedings. IEEE Workshop on, 2001, pp. 141-148.

[10] C. Harris; M. Stephens, "A combined corner and edge detector," in Alvey vision conference, 1988, pp. 50.

[11] J. Shi; C. Tomasi, "Good Features to Track," in Computer Vision and Pattern Recognition, 1994. Proceedings CVPR'94., 1994 IEEE Computer Society Conference on, 1994, pp. 593-600.

[12] D. G. Lowe, "Object Recognition from Local Scale-Invariant Features," in Computer vision, 1999. The proceedings of the seventh IEEE international conference on, 1999, pp. 1150-1157.

[13] H. Bay; T. Tuytelaars; L. V. Gool, "SURF: Speeded Up Robust Features," in Computer Vision–ECCV 2006, 2006, pp. 404-417.

[14] R. Hartley; A. Zisserman, Multiple View Geometry in Computer Vision, 2nd ed. Cambridge University Press, 2003.

[15] S. G. Fowers; D. Lee; G. Xiong, "Improved Library Shelf Reading Using Color Feature Matching of Book-Spine Images," in Control Automation Robotics & Vision (ICARCV), 2010 11th International Conference on, 2010, pp. 2160-2165.

[16] J. Zhao; W. Tao; H. Zhang; K. Yang, "Study on Side Scan Sonar Image Matching Based on the Integration of SURF and Similarity Calculation of Typical Areas," in OCEANS 2010 IEEE-Sydney, 2010, pp. 1-4.

[17] W. Jeong; S. Lee; and Y. Kim, "Statistical Feature Selection Model for Robust 3D Object Recognition," in Advanced Robotics (ICAR), 2011 15th International Conference on, 2011, pp. 402-408.

[18] P. Newman; G. Sibley; M. Smith; M. Cummins; A. Harrison; C. Mei; I. Posner; R. Shade; D. Schroter; L. Murphy; W. Churchill; D. Cole; I. Reid, "Navigating, recognising and describing urban spaces with vision and laser," in International Journal of Robotics Research, Vol. 28, Issue 11-12, pp. 1406-1433, 2009.

[19] 林映辰; 陳佳妍, Large scale 3D scene registration using data from Velodyne LIDAR. National University of Kaohsiung, 2011.

[20] P. J. Besl; N. D. McKay, "A method for registration of 3-d shapes," in IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 14, Issue 2, pp. 586-606, 1992.

[21] R. B. Rusu; N. Blodow; Z. C. Marton; M. Beetz, "Aligning Point Cloud Views using Persistent Feature Histograms," in Intelligent Robots and Systems, 2008. IROS 2008. IEEE/RSJ International Conference on, 2008, pp. 3384-3391.

[22] R. B. Rusu; N. Blodow; M. Beetz; "Fast Point Feature Histograms (FPFH) for 3D Registration," in Robotics and Automation, 2009. ICRA'09. IEEE International Conference on, 2009, pp. 3212-3217.

[23] E. R. Nascimento; W. R. Schwartz, G. L. Oliveira; A. W. Veira; M. F. M. Campos; D. B. Mesquita, "Appearance and Geometry Fusion for Enhanced Dense 3D Alignment," in Graphics, Patterns and Images (SIBGRAPI), 2012 25th SIBGRAPI Conference on, 2012, pp. 47-54.

[24] Z. Zhang, "Flexible Camera Calibration By Viewing a Plane From Unknown Orientations," in Computer Vision, 1999. The Proceedings of the Seventh IEEE International Conference on, 1999, pp. 666-673.

[25] Y. Duan; L. Chen; Y. Wang; M. Yang; X. Qin; S. He; Y. Jia, "A Real-Time System for 3D Recovery of Dynamic Scene with Multiple RGBD Imagers," in Computer Vision and Pattern Recognition Workshops (CVPRW), 2011 IEEE Computer Society Conference on, 2011, pp. 1-8.

[26] S. Suzuki; K. Abe, "Topological structural analysis of digitized binary images by border following," in Computer Vision, Graphics, and Image Processing, 1985, pp. 32-46.

[27] M. Muja; D. G. Lowe, "Fast Approximate Nearest Neighbors with Automatic Algorithm Configuration," in International Conference on Computer Vision Theory and Applications, 2009, pp. 331-340.

[28] P. Henry; M. Krainin; E. Herbst; X. Ren; D. Fox, "RGB-D mapping: Using Kinect-style depth cameras for dense 3D modeling of indoor environments," in International Conference on Computer Vision Theory and Applications (VISSAPP’09), Vol. 31, Issue 5, pp. 647-663, 2012.

[29] 毛紹嘉; 楊傳凱, 3D Model Alignment. National Taiwan University of Science and Technology, 2004.

[30] A. Makadia; A. Patterson; K. Daniilidis, "Fully Automatic Registration of 3D Point Clouds," in International Conference on Computer Vision Theory and Applications (VISSAPP’09), 2006, pp. 1297-1304.

[31] N. Ahmed, "A System for 360° Acquisition and 3D Animation Reconstruction using Multiple RGB-D Cameras," in International Conference on Computer Animation and Social Agents, 2012.

[32] H. Andreasson; T. Stoyanov, "Real Time Registration of RGB-D Data using Local Visual Features and 3D-NDT Registration," in Semantic Perception, Mapping and Exploration Workshop (SPME), 2012, pp. 43.

[33] J. Stuckler; S. Behnke, "Robust Real-Time Registration of RGB-D Images using Multi-Resolution Surfel Representations," in Robotics; Proceedings of ROBOTIK 2012; 7th German Conference on, 2012, pp. 1-4.

[34] I. Dryanovski; C. Jaramillo; J. Xiao, "Incremental Registration of RGB-D Images," in Robotics and Automation (ICRA), 2012 IEEE International Conference on, 2012, pp. 1685-1690.

[35] OpenNI. (2013). OpenNI [Online]. http://www.openni.org/

簡易檢索 / 詳目顯示

相關論文