Responsive image
博碩士論文 etd-0807117-175440 詳細資訊
Title page for etd-0807117-175440
論文名稱
Title
基於模型方法之3D物件於單眼視覺之姿態辨識
Model-based Pose Estimation of 3D Objects Using a Monocular Camera
系所名稱
Department
畢業學年期
Year, semester
語文別
Language
學位類別
Degree
頁數
Number of pages
68
研究生
Author
指導教授
Advisor
召集委員
Convenor
口試委員
Advisory Committee
口試日期
Date of Exam
2017-08-28
繳交日期
Date of Submission
2017-09-11
關鍵字
Keywords
球面座標系、Chamfer matching、針孔相機模型、3D姿態辨識、Homography
Chamfer matching, Pinhole camera model, spherical coordinate, 3D pose recognition, Homography
統計
Statistics
本論文已被瀏覽 5679 次,被下載 815
The thesis/dissertation has been browsed 5679 times, has been downloaded 815 times.
中文摘要
在現在許多工業製程中,如果能辨識待加工或組裝物件的3D姿態,對於製造程序上的自動化將有很大的幫助,但由於3D視覺需要昂貴的感測器以及校正複雜,所以2D視覺方法較受歡迎。因為一般工廠中待加工物件多為金屬或塑膠製品,其表面並無明顯紋理,所以電腦視覺研究中以紋理特徵進行辨識的方法並不適用。於此本論文提出一利用2D單眼視覺系統進行物件辨識的方法。本方法大致上分為兩大部分,第一部分,為離線建立待辨識物件之2D模型,另一為線上模型匹配與姿態辨識。在第一部分建立模型中,首先於Webot模擬軟體中建立待辨識物件之3D模型,在於此環境中加入一台虛擬相機,接著使用此相機圍繞著待辨識物件以球面座標系並且於設定之工作範圍內對待辨識物件拍攝2D影像來建立模型。第二部分,有了2D模型資訊後,使用Chamfer matching對搜尋影像做匹配運算,找出最匹配之模型,之後對此模型與搜尋影像之間計算出Homography矩陣,接著分解(decompose)此矩陣,得到拍攝匹配模型與搜尋影像時相機之間的轉動與移動關係,之後再利用理想針孔相機模型對移動量進行修正,已達到姿態辨識之目的。
Abstract
In today’s many industrial processes, if the 3D pose of a 3D object which will be processed can be recognized, it will be of great help to automate manufacturing processes. The 2D visual methods are more popular than 3D visual methods because 3D vision requires more expensive sensors and it is complicated for correction. Because the general processing of objects in factory are metal or plastic products, their surface is no obvious texture, so the approach rely on texture features does not apply. For above reasons, this paper presents an approach for recognize the 3D pose of a 3D object in a single camera. The proposed approach combines two parts, one is creates 2D models of objects offline and the other is online model matching and pose estimation. In the part of create 2D model, we first create a 3D model of object in the Webot simulation software. In this environment, a virtual camera is added, and then the camera is used to create 2D image model around the object with a spherical coordinate system. After create 2D image model, the chamfer matching was used to match the search image to find the most matching model. Then, the homography matrix is calculated between the most matching model and the search image, and then the matrix is decomposed to get the rotation and translation of camera which between the model and the search image. After that, we use the ideal pinhole camera model to modify the amount of movement to reach the purpose of 3D pose estimation.
目次 Table of Contents
論文審定書 i
致謝 ii
中文摘要 iii
Abstrat iv
目錄 v
圖目錄 vii
表目錄 ix
第一章 緒論 1
1-1 研究動機與目的 1
1-2 相關研究回顧 2
1-3 論文架構 3
第二章 研究背景 4
2-1 CHAMFER MATCHING 4
2-2 角點偵測 5
2-2-1 Harris角點偵測原理 5
2-2-2 Shi-Tomasi角點偵測原理 8
2-3 BRIEF特徵描述子(BINARY ROBUST INDEPENDENT ELEMENTARY FEATURE) 9
2-3-1 BRIEF描述原理 9
2-4 HOMOGRAPHY 矩陣 11
2-4-1 Planar Homography 11
2-4-2 Homography 12
第三章 研究方法 14
3-1 離線建立模型 15
3-2 線上影像匹配 17
3-2-1 Chamfer matching與距離轉換 17
3-2-2 邊緣梯度相似度量測 19
3-2-3 索貝爾算子(Sobel operator) 20
3-3 姿態辨識 21
3-3-1 計算Homography矩陣H 21
3-3-2 二段式角點匹配 23
3-3-3 分解(decompose)Homography矩陣 24
3-3-4 移動量計算 27
第四章 模擬實驗結果 31
4-1 模擬說明 31
4-1-1 建立模型環境 32
4-1-2 驗證環境 33
4-1-3 計算物件以相機座標系描述之3D姿態 34
4-2 模擬實驗結果 38
4-2-1 兩種匹配方法之比較 38
4-2-2 物件不在影像中心之姿態辨識 43
4-2-3 不同θ於姿態辨識之探討 55
第五章 結論與未來展望 56
5-1 結論 56
5-2 未來展望 56
參考文獻 57
參考文獻 References
[1] J. H. M. Byne and J. A. D. W. Anderson , “A CAD based computer vision system ,” Image and Vision Computing, vol. 16, no. 8, pp. 533–539, June 1998.
[2] W. E. L. Grimson and D. P. Huttenlocher, “On the Verification of Hypothesized Matches in Model-Based Recognition,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 13, no. 12, pp. 1201–1213, Dec. 1991.
[3] J. J. Koenderink and A. J. van Doorn, “The Internal Representation of Solid Shape with Respect to Vision,” Biological Cybernetics, vol. 32, pp. 211–216, 1979.
[4] R. D. Schiffenbauer, “A Survey of Aspect Graphs,” Technical Report TR-CIS-2001-01, Dept. of Computer and Information Science, Polytechnic Univ. Brooklyn, New York, Feb. 2001.
[5] S. J. D. A. P. Pentland and A. Rosenfeld, “3-D Shape Recovery Using Distributed Aspect Matching,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 14, no. 2, pp. 174–198, Feb. 1992.
[6] D. W. Eggert, K. W. Bowyer, C. R. Dyer, H. I. Christensen, and D. B. Goldgof, “The Scale Space Aspect Graph,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 15, no. 11, pp. 1114–1130, Nov. 1993.
[7] D. G. Lowe, “Three-Dimensional Object Recognition from Single Two-Dimensional Images,” Artificial Intelligence, vol. 21, no. 3, pp. 335–395, 1987.
[8] M. S. Costa and L. G. Shapiro, “3D Object Recognition and Pose with Relational Indexing,” Computer Vision and Image Understanding, vol. 79, no. 3, pp. 364–407, Sept. 2000.
[9] I. Weiss and M. Ray, “Model-Based Recognition of 3D Objects from Single Images,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 23, no. 2, pp. 116–128, Feb. 2001.
[10] P. Han, G. Zhao, “Cad-based 3d objects recognition in monocular images for mobile augmented reality”, Computers & Graphics, vol. 50, pp. 36–46, 2015.
[11] S. Belongie, J. Malik, and J. Puzicha, “Shape Matching and Object Recognition Using Shape Contexts,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 24, no. 4, pp. 509–522, Apr. 2002.
[12] H. Bay, T. Tuytelaars, and L. Van Gool, “SURF: Speeded Up Robust Features,” in Proc. Eur. Conf. Comput. Vision, Graz, Austria, May 2006, pp. 404–417.
[13] H. G. Barrow, J. M. Tenenbaum, R. C. Bolles, and H. C. Wolf, “Parametric correspondence and chamfer matching: Two new techniques for image matching,” in Proc. 5th Int. Joint Conf. Artif. Intell., 1977, pp. 659–663.
[14] C. Harris and M. Stephens, “A combined corner and edge detector,” in Proc. Alvey Vis. Conf., 1988, pp. 147–151.
[15] H. P. Moravec, “Toward automatic visual obstacle avoidance,” in Proc. 5th Int. Joint Conf. Artif. Intell., Cambridge, MA, USA, 1977, p. 584.
[16] M. Calonder, V. Lepetit, C. Strecha, and P. Fua, “BRIEF: Binary robust independent elementary features,” in Proc. 11th Eur. Conf. Cpmput. Vis., 2010, pp. 778–792.
[17] J. Canny, “A computational approach to edge detection,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. PAMI-8, no.6, pp. 679–698. Nov. 1986.
[18] C. Steger, “Occlusion, Clutter, and Illumination Invariant Object Recognition,” Int'l Archives of Photogrammetry and Remote Sensing, vol. 34, part 3A, pp. 345–350, 2002.
[19] G. Borgefores, “Hierarchical chamfer matching: A parametric edge matching algorithm”, IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 10, no. 6, pp. 849–856, Nov. 1988.
[20] E. Malis and M. Vargas, “Deeper understanding of the homography decomposition for vision-based control,” INRIA, Sophia Antipolis, France, Tech. Rep. RR-6303, 2007.
電子全文 Fulltext
本電子全文僅授權使用者為學術研究之目的,進行個人非營利性質之檢索、閱讀、列印。請遵守中華民國著作權法之相關規定,切勿任意重製、散佈、改作、轉貼、播送,以免觸法。
論文使用權限 Thesis access permission:校內校外完全公開 unrestricted
開放時間 Available:
校內 Campus: 已公開 available
校外 Off-campus: 已公開 available


紙本論文 Printed copies
紙本論文的公開資訊在102學年度以後相對較為完整。如果需要查詢101學年度以前的紙本論文公開資訊,請聯繫圖資處紙本論文服務櫃台。如有不便之處敬請見諒。
開放時間 available 已公開 available

QR Code