水下拖曳式攝影系統(Towed Camera System; TowCam) 在施測時，往往會受到
海流影響造成載具艏向(Heading) 與運動方向不同向，存在一艏向偏角α。這個角
度會使得載具上搭載的照相機所拍攝之影像在日後進行影像拼接(Photo mosaicking)
時，所拼接出來的海底影像呈鋸齒狀，造成圖片可利用的完整區域較少。為了探
討此問題，本研究使用加速強健特徵(Speeded-Up Robust Features; SURF) 演算
法萃取影像特徵點，並以特徵點為基礎建立一載具艏向偏角估算方法，以利後
續應用於水下拖曳式攝影系統上視覺伺服控制之開發。我們亦使用TP-TowCam
(Twist-Pair TowCam) 所拍攝之實海域影像對估算方法進行測試，探討了SURF 演
算法中Hessian 閥值於本研究所提出之演算法的調整策略。而我們對演算法測試的
結果顯示，在拍攝環境較差的海底下，仍然能夠提供不錯的估算結果，且單位時間
內可容許的位移量也相對較大。

When TowCam (Towed Camera System) is under the deep sea exploration, its
moving track and heading always have different direction with degree of  α caused
by the current. This situation will make images shot by the camera mounted on
the vehicle to difficultly be processed by the photo mosaicking method and get the
deformed or unaccepted mosaic results. In this thesis, we used the SURF (speeded-up
robust feature) algorithm to extract the critical point of the shot image to estimate
the declination between the heading and the track for future development of visual
servo control of TowCam. The images shot by the Twist-Pair TowCam (TP-TowCam)
were also used to check the reliable of estimated algorithm in the filed experiment test
and to find the best Hessian threshold value of SURF algorithm for the estimation.
The results of algorithm showed the good estimation even in the turbid condition and
the acceptable displacement of moving vehicle in the unit time was also increased,
indicating the improvement of evaluation.

目錄
1 緒論 1
1.1 前言. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 研究動機與目的. . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 文獻回顧. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4 論文架構. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2 影像處理 18
2.1 OpenCV 影像函式庫. . . . . . . . . . . . . . . . . . . . . . . .19
2.2 灰階轉換. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3 影像解析度降階. . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.4 影像強化. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.5 SURF 特徵萃取. . . . . . . . . . . . . . . . . . . . . . . . . . . .23
2.6 特徵匹配. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3 載具運動艏向偏角估算 35
3.1 前言. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2 載具運動評估. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.3 載具運動估算. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4 影像分析 48
4.1 Hessian 閥值對估算之影響. . . . . . . . . . . . . . . . . . 48
4.2 影像解析度降階與強化對估算之影響. . . . . . . . . 54
4.3 影像時間間距與估算方法分析. . . . . . . . . . . . . . . 66
5 討論與結論 76
5.1 討論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.2 結論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.3 建議. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
參考文獻 82
附錄A SURF 演算法 87

[1] 鐘三雄、林曉武、林裕程、王詠絢、陳松春、楊燦堯、劉家瑄、陳柏淳、費立
沅，台灣西南海域深海數位照相探測初步成果，臺灣地球科學聯合學術研討會，
ER-P-02，2007。
[2] 黃國銘、林曉武、林裕程，底拖式照相系統應用在台灣深海探測，第33 屆海洋
工程研討會論文集，pp. 881-885，2011。
[3] R. Prados, R. Garcia, N. Gracias, J. Escartin, and L. Neumann, “A novel blending
technique for underwater gigamosaicing,” IEEE Journal of Oceanic Engineering,
vol. 37, pp. 626-644, 2012.
[4] S. van der Zwaan, A. Bernardino, and J. Santos-Victor, “Visual station keeping for
floating robots in unstructured environments,” Robotics and Autonomous System,
vol. 39, pp. 145-155, 2002.
[5] S. Negahdaripour, and P. Firoozfam, “An ROV stereovision system for shiphull
inspection,” IEEE Journal of Oceanic Engineering, vol. 31, pp. 551–564, 2006.
[6] A. Ortiz, J. Antich, and G. Oliver, “A particle filter-based approach for tracking
undersea narrow telecommunication cables,” Machine Vision and Applications,
vol. 22, pp. 283–302, 2008.
[7] P. Drews, V. Kuhn, and S. Gomes, “Tracking system for underwater inspection
using computer vision,” Offshore and Marine Technology: Science and Innovation
(NAVTEC), 2012 International Conference on. IEEE, pp. 27-30, 2012.
[8] M. Dunbabin, K. Usher, and P. Corke, “Visual motion estimation for an autonomous
underwater reef monitoring robot,” Field and Service Robotics, Springer
Tracts in Advanced Robotics vol. 25, pp. 31-42, 2006.
[9] R. M. Eustice, O. Pizarro, and H. Singh, “Visually augmented navigation for
autonomous underwater vehicles,” Oceanic Engineering, IEEE Journal of, vol. 33,
pp. 103-122, 2008.
[10] N. R. Gracias, S. van der Zwaan, A. Bernardino, and J. Santos-Victor, “Mosaicbased
navigation for autonomous underwater vehicles,” IEEE Journal of Oceanic
Engineering, vol. 28, pp. 609-624, 2003.
[11] S. Negahdaripour, X. Xu, and L. Jin, “Direct estimation of motion from sea floor
images for automatic station-keeping of submersible platforms,” IEEE Journal of
Oceanic Engineering, vol. 24, 1999.
[12] J. Canny, “A computational approach to edge detection,” IEEE transaction on
pattern analysis and machine intelligence, vol. 8, pp. 679–698, 1986.
[13] C. Harris, and M. Stephens, “A combined corner and edge detector,” Proceedings
of the 4th alvey vision conference, pp. 147–151, 1988.
[14] H. P. Moravec, “Towards automatic visual obstacle avoidance,” Proceedings of the
5th international joint conference on Artificial intelligence, 1977.
[15] Jan J. Koenderink, “The structure of images,” Biological cybernetics, vol. 50, pp.
363-370, 1984.
[16] Luc M.J Florack, Bart M. ter Harr Romeny, Jan J. Koenderink, and Max A.
Viergever, “Scale and the differential structure of images,” Image and Vision Computing,
vol. 10, pp. 376-388, 1992.
[17] T. Lindeberg, Scale-space theory in computer vision, Kluwer Academic Springer,
1994.
[18] D. G. Lowe, “Object recognition from local scale-invariant features,” International
Conference of Computer Vision, vol. 60, pp. 1150-1157, 1999.
[19] D. G. Lowe, “Distinctive image features from scale-invariant keypoints,” International
Journal of Computer Vision, vol. 60, pp. 91-110, 2004.
[20] H. Bay, T. Tuytelaars, and L.V. Gool, “SURF: Speeded Up Robust Features,”
Proceedings of the 9th European Conference on Computer Vision, pp. 404-417,
2006.
[21] H. Bay, A. Ess, T. Tuytelaars, and L.V. Gool, “Speeded-Up Robust Features
(SURF),” Computer Vision and Image Understanding, vol. 110, pp. 346-359, 2008.
[22] L. Juan, O. Gwun, “A comparison of SIFT, PCA-SIFT and SURF,” International
Journal of Image Processing, vol. 3, pp. 143-152, 2009.
[23] K. Yan, and R. Sukthankar, “PCA-SIFT: A more distinctive representation for
local image descriptors,” IEEE Computer Society Conference on Computer Vision
and Pattern Recognition (CVPR’04), vol. 2, pp. 506-513, 2004.
[24] OpenCV, http://opencv.org/, 2013.
[25] G. Bradski, and K. Adrian, Learning OpenCV: Computer Vision with the OpenCV
Library, O’Reilly Media, Inc, 1st edition, 2008.
[26] 賴岱佑、劉敏，數位影像處理技術手冊，初版，文魁資訊股份有限公司，2007。
[27] R. C. Gonzalez, and R. E. Woods, Digital Image Processing, 2rd. Prentice Hall,
2002.
[28] 鍾國亮，影像處理與電腦視覺，第五版，東華書局，2012。
[29] P. Viola, and M. Jones, “Rapid object detection using a boosted cascade of simple
features,” IEEE Computer Society Conference on Computer Vision and Pattern
Recognition (CVPR’01), vol. 1, pp. I-511-I-518. 2001.
[30] 王永明、王貴錦，圖像局部不變性特徵與描述，第一版，國防工業出版社，
2010。
[31] C. Evans, Notes on the opensurf library, University of Bristol, Tech. Rep. CSTR-
09-001, 2009.
[32] P. Smith, D. Sinclair, R. Cipolla, and K. Wood, “Effective corner matching,”
British Machine Vision Conference, pp. 1-12, 1998.
[33] 李振偉，水下無人載具之視覺懸停控制，國立中山大學機械與機電工程學系碩
士論文，2008。
[34] 呂璨朱，水下無人載具之免校正視覺伺服控制，國立中山大學機械與機電工程
學系碩士論文，2010。