臺灣博碩士論文加值系統

無人飛行載具(UAV)的應用已廣泛於軍事、科技與民生用途。當UAV遭遇緊急狀況需進行迫降時，UAV必須有能力自行搜尋一塊平坦的區域進行迫降。搜尋的方式可以透過雷達、雷射、超音波或影像辨識方式進行，而本論文是完全採用立體視覺的方式來進行地形平坦度分析。本論文是透過裝設在UAV機體下方的相機得到地面之即時影像，並參考紅外線測距儀或高度計以取得目前飛行高度，再利用NCC (Normalized Cross Correlation)匹配演算法先後對邊緣與非邊緣像素點進行匹配計算得到影像的真實深度。藉由此深度資訊來判斷地形平坦度，如此UAV將能自行搜尋出一塊可進行迫降的區域。

Unmanned Aerial Vehicle (UAV) has been widely applied in military, technology and the people’s livelihood. When UAV encountering some situations needs to do emergency landing, it should have the ability to search a flat area for landing. The search ways can use radar, laser, ultrasound, or image recognition, and this paper completely uses the stereo vision approach to analysis flatness. This paper uses a pair of cameras, which are deployed under the body of UAV to get the real-time images of ground, refers infrared distance meter or altimeter to evaluate the current altitude, and then uses NCC algorithm to find edge and non-edge matched pixels to compute the real depths of the terrain. Based on the obtained depth information, UAV can autonomously determine the terrain flatness and find an area for emergency landing.

摘要 i
英文摘要 ii
誌謝 iii
目錄 iv
圖目錄 vi
第一章 緒論 1
1-1 研究動機 1
1-2 主要貢獻 1
1-3 章節架構 1
第二章 相關背景知識與定理推導 3
2-1 無人飛行載具 3
2-1-1 無人飛行載具的歷史 3
2-1-2 無人飛行載具的發展 4
2-2 四旋翼直昇機 4
2-3 緊急降落系統 6
2-4 三維掃描技術 6
2-5 深度估算技術 7
2-5-1 匹配代價估算 7
2-5-2 匹配代價聚合與視差估算/優化 7
2-5-3 視差細化 8
第三章 深度計算與平坦度分析 9
3-1 雙邊濾波器 10
3-2 坎尼邊緣偵測 11
3-3 邊緣匹配 12
3-4 非邊緣匹配 14
3-5 深度計算 15
3-6 深度補償 16
3-7 深度單位轉換 19
3-8 平坦度分析 20
3-9 降落範圍 20
3-9-1 範圍大小計算 20
3-9-2 平坦度影像 22
第四章 實驗模擬與分析 24
4-1 實驗環境 24
4-2 深度影像模擬 25
4-2-1 針對邊緣匹配進行檢測 25
4-2-2 深度補償機制之模擬 27
4-2-3 真實深度表之模擬 29
4-3 真實影像之模擬 30
第五章 結論與展望 33
參考文獻 34

[1] 「無人航空載具」，http://zh.wikipedia.org/wiki/無人航空載具，2015。
[2] 「UAV無人飛行載具之應用」，http://www.shs.edu.tw/works/essay/2007/10/
2007102919131097.pdf，2007。
[3] 「四軸飛行器」，http://zh.wikipedia.org/wiki/四軸飛行器#cite_ref-1，2015。
[4] 「多軸飛行器」，http://zh.wikipedia.org/wiki/多軸飛行器，2015。
[5] 鄭文昌和王偉誠，「四旋翼 UAV定點懸置之模糊控制器設計與模擬」，資訊科技國際期刊，第八卷，第二期，2014。
[6] L. Mejias and D. Fitzgerald, “A Multi-Layered Approach for Site Detection in UAS Emergency Landing Scenarios Using Geometry-based Image Segmentation,” in Proc. International Conference on Unmanned Aircraft Systems, pp. 366-372, May 2013.
[7] 「3D掃描器」，http://zh.wikipedia.org/wiki/三維掃描儀，2015。
[8] 陳育菘、楊志弘和廖育昇，「以SGM為基礎之三維深度影像估算的研究」，中國機械工程學會第二十九屆全國學術研討會論文集，編號1438，2012。
[9] D. Scharstein, R. Szeliski, and R. Zabih, “A Taxonomy and Evaluation of Dense Two-Frame Stereo Correspondence Algorithms,” in Proc. IEEE Workshop on Stereo and Multi-Baseline Vision, pp. 131-140, Dec. 2001.
[10] “Correlation Based Similarity Measures-Summary,” https://siddhantahuja.word-
press.com/2010/04/11/correlation-based-similarity-measures-summary/, 2010.
[11] “Hamming Distance,” http://en.wikipedia.org/wiki/Hamming_distance, 2015.
[12] R. Szeliski and D. Scharstein, “Sampling the Disparity Space Image,” IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 419-425, Mar. 2004.
[13] H. Hirschmuller and D. Scharstein, “Evaluation of Cost Functions for Stereo Matching,” in Proc. IEEE Conference on Computer Vision and Pattern Recognition, pp. 1-8, June 2007.
[14] “Belief Propagation,” http://en.wikipedia.org/wiki/Belief_propagation, 2015.
[15] 「置信度傳播算法」，http://blog.csdn.net/aspirinvagrant/article/details/40862237，2014。
[16] “Dynamic Programming,” http://www.csie.ntnu.edu.tw/~u91029/DynamicProgra
mming.html, 2015.
[17] “Maximum Flow Problem,” http://en.wikipedia.org/wiki/Maximum_flow_
problem, 2015.
[18] Y. Y. Boykov and M.-P. Jolly, “Interactive Graph Cuts for Optimal Boundary & Region Segmentation of Objects in N-D Images,” in Proc. IEEE International Conference on Computer Vision, vol. 1, pp. 105-112, July 2001.
[19] H. Hirschmuller, “Accurate and Efficient Stereo Processing by Semi-Global Matching and Mutual Information,” in Proc. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 2, pp. 807-814, June 2005.
[20] 「雙邊濾波器」，http://zh.wikipedia.org/wiki/雙邊濾波器，2013。
[21] C. Tomasia and R. Manduchi, “Bilateral Filtering for Gray and Color Images,” in Proc. 6th IEEE International Conference on Computer Vision, pp. 839-846, Jan. 1998.
[22] V. Mall, A.-K. Roy, and S.-K. Mitra, “Detection of Structural Tampering in a Digital Image Using Canny Edge Detector,” in Proc.International Conference on Informatics, Electronics & Vision, pp. 1-7, May 2013.
[23] J. Canny, “A Computational Approach to Edge Detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. PAMI-8, no. 6, pp. 679-698, Nov. 1986.
[24] “Gaussian Filter,” http://en.wikipedia.org/wiki/Gaussian_filter, 2015.
[25] 「Canny邊緣檢測算法原理及其VC實現詳解(一)」，http://blog.csdn.net/
likezhaobin/article/details/6892176，2011。
[26] 「Canny邊緣檢測基本原理與C++實現」，http://rockyhotshow.blogspot.tw/
2011/10/cannyc.html，2011。
[27] L.-D. Stefano, S. Mattoccia, and M. Mola, “An Efficient Algorithm for Exhaustive Template Matching based on Normalized Cross Correlation,” in Proc. 12th International Conference on Image Analysis and Processing, pp. 322-327, Sep. 2003.
[28] R. Jain, R. Kasturi, and B. G. Schunck, Machine Vision, McGraw-Hill, New York, 1995.