Publications-Theses

Article View/Open

Publication Export

Google ScholarTM

NCCU Library

Citation Infomation

Related Publications in TAIR

題名 使用光束調整法與多張影像做相機效正與三維模型重建
Using bundle adjustment for camera Calibration and 3D reconstruction from multiple images
作者 蔡政君
Tsai, Jeng Jiun
貢獻者 何瑁鎧
Hor, Maw Kae
蔡政君
Tsai, Jeng Jiun
關鍵詞 影像處理
極限幾何
對應點
補綴面
再投影誤差
image processing
epipolar geometry
corresponding points
sparse bundle adjustment
patch
normalized cross correlation
reprojection error
日期 2009
上傳時間 9-May-2016 15:29:03 (UTC+8)
摘要 自動化三維建模需要準確的三維點座標,而三維點的位置則依賴高精度的對應點,因此對應點的尋找一直是此領域的研究議題,而使用稀疏光束調整法(SBA:Sparse Bundle Adjustment)來優化相機參數也是常用的作法,然而若三維點當中有少數幾個誤差較大的點,則稀疏光束調整法會受到很大的影響。我們採用多視角影像做依據,找出對應點座標及幾何關係,在改善對應點位置的步驟中,我們藉由位移三維點法向量來取得各種不同位置的三維補綴面(3D patch),並根據投影到影像上之補綴面的正規化相關匹配法(NCC:Normalized Cross Correlation)來改善對應點位置。利用這些改善過的點資訊,我們使用稀疏光束調整法來針對相機校正做進一步的優化,為了避免誤差較大的三維點影響到稀疏光束調整法的結果,我們使用穩健的計算方法來過濾這些三維點,藉由此方法來減少再投影誤差(reprojection error),最後產生較精準的相機參數,使用此參數我們可以自動化建出外型架構較接近真實物體的模型。
Automated 3D modeling of the need for accurate 3D points, and location of the 3D points depends on the accuracy of corresponding points, so the search for corresponding points in this area has been a research topic, and the use of SBA(Sparse Bundle Adjustment) to optimize the camera parameters is also a common practice, however, if there are a few more error 3D points, the SBA will be greatly affected. In this paper, we establish the corresponding points and their geometry relationship from multi-view images. And the 3D patches are used to refine point positions. We translate the normal to get many patches, and project them into visible images. The NCC(Normalized Cross Correlation) values between patches in reference image and patches in visible image are used to estimate the best correspondence points. And they are used to get better camera parameters by SBA(sparse bundle adjustment). Furthermore, it is because that it usually exist outliers in the data observed, and they will influence the results by using SBA. So, we use our robust estimation method to resist the outliers. In our experiment, SBA is used to filter some outliers to reduce the reprojection error. After getting more precise camera parameters, we use them to reconstruct the 3D model more realistic.
參考文獻 [1] Al-Hanbali, N. and B. Sadoun, “3D GIS Modeling of BAU: Planning Prospective and Implementation Aspects”, IEEE/ACS International Conference on Computer Systems and Applications, pp.566-571, 2007.
     [2] Bouguet, J.-Y., Camera calibration toolbox for matlab.
     [3] Furukawa, Y. and J. Ponce, “Accurate Camera Calibration from Multi-View Stereo and Bundle Adjustment”, IEEE Conference on Computer Vision and Pattern Recognition, pp.1-8, 2008.
     [4] Furukawa, Y. and J. Ponce, “Accurate, Dense, and Robust Multi-View Stereopsis”, IEEE Conference on Computer Vision and Pattern Recognition, 2007.
     [5] Furukawa Y. and J. Ponce, PMVS (http://wwwcvr.ai.uiuc.edu/~yfurukaw/research/pmvs).
     [6] Hartley, R. I. and A. Zisserman, Multiple View Geometry in Computer Vision, Cambridge University Press, 2003.
     [7] Lourakis, M. and A. Argyros, SBA: A Generic Sparse Bundle Adjustment C/C++ Package Based on the Levenberg- Marquardt Algorithm (http://www.ics.forth.gr/~lourakis/sba/).
     [8] Martinec, D. and T. Pajdla, “Robust Rotation and Translation Estimation in Multiview Reconstruction”, IEEE conference on Computer Vision and Pattern Recognition, 2007.
     [9] Paolo, C., C. Marco, C. Massimiliano, G. Fabio, and R. Guido, MeshLab(http://meshlab.sourceforge.net/).
     [10] Seitz, S. M., B. Curless, J. Diebel, D. Scharstein, and R. Szeliski, Multi-View Stereo Evaluation(http://vision.middlebury.edu/mview/).
     [11] Tang, C.-Y., H.-L. Chou, Y.-L. Wu, and Y.-H. Ding, “Robust Fundamental Matrix Estimation Using Coplanar Constraints”, International Journal of Pattern Recognition and Artificial Intelligence, pp.783-805, 2008.
     [12] Tang, C.-Y., Y.-L. Wu, and Y.-H. Lai, “Fundamental Matrix Estimation Using Evolutionary Algorithms with Multi-Objective Functions”, Journal of Information Science and Engineering, pp.785-800, 2008.
     [13] Tang, C.-Y., Y.-L. Wu, Maw-Kae Hor, and Wen-Hung Wang, “Modified SIFT Descriptor for Image Matching under Interference”, International Conf. Machine Learning and Cybernetics, pp.3294-3300, 2008.
     [14] 賴易進,"由地圖建構城市三維模型",國立政治大學資訊科學系碩士論文,台北,民國95年11月。
     [15] 詹凱軒,"由地面光達資料自動重建建物模型之研究",國立政治大學資訊科學系碩士論文,台北,民國96年7月。
描述 碩士
國立政治大學
資訊科學學系
96753002
資料來源 http://thesis.lib.nccu.edu.tw/record/#G0096753002
資料類型 thesis
dc.contributor.advisor 何瑁鎧zh_TW
dc.contributor.advisor Hor, Maw Kaeen_US
dc.contributor.author (Authors) 蔡政君zh_TW
dc.contributor.author (Authors) Tsai, Jeng Jiunen_US
dc.creator (作者) 蔡政君zh_TW
dc.creator (作者) Tsai, Jeng Jiunen_US
dc.date (日期) 2009en_US
dc.date.accessioned 9-May-2016 15:29:03 (UTC+8)-
dc.date.available 9-May-2016 15:29:03 (UTC+8)-
dc.date.issued (上傳時間) 9-May-2016 15:29:03 (UTC+8)-
dc.identifier (Other Identifiers) G0096753002en_US
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/95265-
dc.description (描述) 碩士zh_TW
dc.description (描述) 國立政治大學zh_TW
dc.description (描述) 資訊科學學系zh_TW
dc.description (描述) 96753002zh_TW
dc.description.abstract (摘要) 自動化三維建模需要準確的三維點座標,而三維點的位置則依賴高精度的對應點,因此對應點的尋找一直是此領域的研究議題,而使用稀疏光束調整法(SBA:Sparse Bundle Adjustment)來優化相機參數也是常用的作法,然而若三維點當中有少數幾個誤差較大的點,則稀疏光束調整法會受到很大的影響。我們採用多視角影像做依據,找出對應點座標及幾何關係,在改善對應點位置的步驟中,我們藉由位移三維點法向量來取得各種不同位置的三維補綴面(3D patch),並根據投影到影像上之補綴面的正規化相關匹配法(NCC:Normalized Cross Correlation)來改善對應點位置。利用這些改善過的點資訊,我們使用稀疏光束調整法來針對相機校正做進一步的優化,為了避免誤差較大的三維點影響到稀疏光束調整法的結果,我們使用穩健的計算方法來過濾這些三維點,藉由此方法來減少再投影誤差(reprojection error),最後產生較精準的相機參數,使用此參數我們可以自動化建出外型架構較接近真實物體的模型。zh_TW
dc.description.abstract (摘要) Automated 3D modeling of the need for accurate 3D points, and location of the 3D points depends on the accuracy of corresponding points, so the search for corresponding points in this area has been a research topic, and the use of SBA(Sparse Bundle Adjustment) to optimize the camera parameters is also a common practice, however, if there are a few more error 3D points, the SBA will be greatly affected. In this paper, we establish the corresponding points and their geometry relationship from multi-view images. And the 3D patches are used to refine point positions. We translate the normal to get many patches, and project them into visible images. The NCC(Normalized Cross Correlation) values between patches in reference image and patches in visible image are used to estimate the best correspondence points. And they are used to get better camera parameters by SBA(sparse bundle adjustment). Furthermore, it is because that it usually exist outliers in the data observed, and they will influence the results by using SBA. So, we use our robust estimation method to resist the outliers. In our experiment, SBA is used to filter some outliers to reduce the reprojection error. After getting more precise camera parameters, we use them to reconstruct the 3D model more realistic.en_US
dc.description.tableofcontents 第一章 緒論.................................................1
     1.1 研究動機與目的..........................................1
     1.2 問題描述................................................3
     1.3 系統架構與流程說明.......................................4
     1.4 本論文的貢獻............................................5
     1.5 論文章節架構............................................6
     第二章 相關研究.............................................8
     第三章 背景知識............................................15
     3.1 極線幾何與投影幾何......................................15
     3.2 多視角影像.............................................17
     3.3 正規化相關匹配法........................................18
     3.4 再投影誤差.............................................19
     第四章 影像對應初始化.......................................21
     4.1 系統架構...............................................21
     4.2 對應點座標初始化........................................22
     4.3 隨機取樣...............................................25
     第五章 對應點位置改善.......................................26
     5.1 補綴面掃描法...........................................27
     5.2 法向量旋轉與平移法......................................31
     5.2.1 以法向量改善對應點位置................................31
     5.2.2 點座標過濾...........................................35
     5.3 法向量平移法...........................................36
     5.4 三維點位置決定.........................................37
     第六章 相機參數之調整.......................................40
     6.1 稀疏光束調整法.........................................40
     6.2 稀疏光束調整法前處理....................................41
     第七章 實驗結果............................................44
     7.1 再投影誤差期望值之定義..................................45
     7.2 隨機取樣..............................................46
     7.3 對應點位置改善.........................................49
     7.4 點座標過濾.............................................50
     7.5 稀疏光束調整法前處理....................................52
     7.6 建模..................................................58
     第八章 結論................................................62
     8.1 結論..................................................62
     8.2 未來發展...............................................63
     參考文獻...................................................65
zh_TW
dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0096753002en_US
dc.subject (關鍵詞) 影像處理zh_TW
dc.subject (關鍵詞) 極限幾何zh_TW
dc.subject (關鍵詞) 對應點zh_TW
dc.subject (關鍵詞) 補綴面zh_TW
dc.subject (關鍵詞) 再投影誤差zh_TW
dc.subject (關鍵詞) image processingen_US
dc.subject (關鍵詞) epipolar geometryen_US
dc.subject (關鍵詞) corresponding pointsen_US
dc.subject (關鍵詞) sparse bundle adjustmenten_US
dc.subject (關鍵詞) patchen_US
dc.subject (關鍵詞) normalized cross correlationen_US
dc.subject (關鍵詞) reprojection erroren_US
dc.title (題名) 使用光束調整法與多張影像做相機效正與三維模型重建zh_TW
dc.title (題名) Using bundle adjustment for camera Calibration and 3D reconstruction from multiple imagesen_US
dc.type (資料類型) thesisen_US
dc.relation.reference (參考文獻) [1] Al-Hanbali, N. and B. Sadoun, “3D GIS Modeling of BAU: Planning Prospective and Implementation Aspects”, IEEE/ACS International Conference on Computer Systems and Applications, pp.566-571, 2007.
     [2] Bouguet, J.-Y., Camera calibration toolbox for matlab.
     [3] Furukawa, Y. and J. Ponce, “Accurate Camera Calibration from Multi-View Stereo and Bundle Adjustment”, IEEE Conference on Computer Vision and Pattern Recognition, pp.1-8, 2008.
     [4] Furukawa, Y. and J. Ponce, “Accurate, Dense, and Robust Multi-View Stereopsis”, IEEE Conference on Computer Vision and Pattern Recognition, 2007.
     [5] Furukawa Y. and J. Ponce, PMVS (http://wwwcvr.ai.uiuc.edu/~yfurukaw/research/pmvs).
     [6] Hartley, R. I. and A. Zisserman, Multiple View Geometry in Computer Vision, Cambridge University Press, 2003.
     [7] Lourakis, M. and A. Argyros, SBA: A Generic Sparse Bundle Adjustment C/C++ Package Based on the Levenberg- Marquardt Algorithm (http://www.ics.forth.gr/~lourakis/sba/).
     [8] Martinec, D. and T. Pajdla, “Robust Rotation and Translation Estimation in Multiview Reconstruction”, IEEE conference on Computer Vision and Pattern Recognition, 2007.
     [9] Paolo, C., C. Marco, C. Massimiliano, G. Fabio, and R. Guido, MeshLab(http://meshlab.sourceforge.net/).
     [10] Seitz, S. M., B. Curless, J. Diebel, D. Scharstein, and R. Szeliski, Multi-View Stereo Evaluation(http://vision.middlebury.edu/mview/).
     [11] Tang, C.-Y., H.-L. Chou, Y.-L. Wu, and Y.-H. Ding, “Robust Fundamental Matrix Estimation Using Coplanar Constraints”, International Journal of Pattern Recognition and Artificial Intelligence, pp.783-805, 2008.
     [12] Tang, C.-Y., Y.-L. Wu, and Y.-H. Lai, “Fundamental Matrix Estimation Using Evolutionary Algorithms with Multi-Objective Functions”, Journal of Information Science and Engineering, pp.785-800, 2008.
     [13] Tang, C.-Y., Y.-L. Wu, Maw-Kae Hor, and Wen-Hung Wang, “Modified SIFT Descriptor for Image Matching under Interference”, International Conf. Machine Learning and Cybernetics, pp.3294-3300, 2008.
     [14] 賴易進,"由地圖建構城市三維模型",國立政治大學資訊科學系碩士論文,台北,民國95年11月。
     [15] 詹凱軒,"由地面光達資料自動重建建物模型之研究",國立政治大學資訊科學系碩士論文,台北,民國96年7月。
zh_TW