簡易檢索 / 詳目顯示

研究生: 張詠惠
Yung-Hui Chang
論文名稱: 基於立體影像匹配的深度量測演算法
A Depth Measurement Algorithm Based on Stereo Matching
指導教授: 王乃堅
Nai-Jian Wang
口試委員: 黃騰毅
none
蔡超人
none
姚嘉瑜
none
郭景明
none
學位類別: 碩士
Master
系所名稱: 電資學院 - 電機工程系
Department of Electrical Engineering
論文出版年: 2012
畢業學年度: 100
語文別: 中文
論文頁數: 66
中文關鍵詞: 立體視覺雙眼視覺影像校正自動調整樣版大小均值移動演算法
外文關鍵詞: Stereo vision, Binocular vision, Image calibration, Adaptive Bandwidth Mean Shift algorithm
相關次數: 點閱:241下載:6
分享至:
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
  • 本論文的主要目標為建立一個立體視覺系統以估測目標點與攝影機的距離。單張影像無法提供確切的深度資訊,必須藉由多張影像已達成此目的。在立體視覺中尋找對應點是非常重要的。一般尋找對應點會針對固定目標點的特色做處理,例如:顏色、形狀等,但是如果有這樣的限制,立體測距將會被特定的目標點限制,而降低其實用性。
    本論文可以分為三大步驟,第一步驟利用兩個一般的網路攝影機擷取畫面,由於我們必須解決徑向扭曲和雙攝影機光軸不平行的問題,因此需要做影像校正。第二步驟依據目標點附近的紋理選擇尋找對應點的方法,若目標點附近紋理豐富,則利用紋理使用絕對差值總和演算法,若相反,則利用顏色分佈使用自動調整樣版大小均值移動演算法。第三步驟利用三角成像原理,計算出目標點與攝影機之間的距離。
    在我們的實驗結果中顯示當我們對影像校正做進一步的修改後,我們可以利用低價位的攝影機,和依據目標點的特性來估測其距離,並且擁有不錯的準確度其最大平均誤差在3.04%以內。


    The purpose of this thesis is to build a stereo vision system for estimating distances between the target point and the webcams. We construct the stereo system in order to estimate exact information of depth by locating the corresponding points. In general, locating the corresponding point is achieved by the features of the object, such as color and shape. However, the selection of target points will be limited and the adaptability of the system will be declined. Thus, the method was proposed to locate the corresponding point according to the texture around the target point.
    This system consists of three major steps. First, Images are captured by two webcams. The image calibration is needed in order to get the parameters of the webcams, to deal with the distortion, and to compensate the unparalleled optical axes. Second, two different ways are used to find the corresponding point according to the texture around the target point. If the texture around the target point is complex, we use the Sum of Absolute Difference algorithm to find the corresponding point. Otherwise, the Adaptive Bandwidth Mean Shift algorithm is used. Finally, by the triangular perspective projection, the distance between the target point and webcams can be estimated.
    After the proposed image calibration algorithm, the system can be implemented with low-price cameras. The experimental results show that the accuracy is well with mean error rate less than 3.04%.

    摘要 I ABSTRACT II 誌謝 III 目錄 IV 圖目錄 VI 表目錄 VIII 第一章 緒論 1 1.1 研究目的與動機 1 1.2 相關研究回顧 1 1.2.1 攝影機架設 1 1.2.2 特徵點匹配 2 1.2.3 距離估測 3 1.3 論文組織 3 第二章 系統架構與實驗環境 5 2.1 開發環境與介面 5 2.2 系統流程 7 第三章 雙眼視覺測距 9 3.1 影像校正 9 3.1.1 內部參數 10 3.1.2 外部參數 12 3.1.3 參數校正 12 3.1.4 修正旋轉矩陣 19 3.2 亮度矯正 21 3.3 基於自動調整樣版大小均值移動之比對 22 3.3.1 色彩模型轉換 22 3.3.2 核心函數 24 3.3.3 目標點之色彩分佈模型 28 3.3.4 候選點之色彩分佈模型 30 3.3.5 相似度計算 32 3.3.6 平均值位移演算法 35 3.4 絕對差值總和之比對 41 3.5 比對方式的選擇 42 3.6 距離估測 44 第四章 實驗結果 46 4.1 影像校正參數 46 4.2 比對方式的選擇 47 4.3 誤差分析 50 4.4 系統所需時間 53 4.5 測量數據 53 第五章 結論與未來展望 55 5.1 結論 55 5.2 未來研究方向 55 參考文獻 56 作者簡介 58

    [1] J. S. Lee, C. W. Seo, and E. S. Kim, "Implementation of opto-digital stereo object tracking system," Optics Communications, Vol. 200, pp. 73-85, 2001.

    [2] 徐瑋甯,"立體視覺人形機器人於競走項目之實現," 淡江大學電機工程學系碩士論文,2011。

    [3] D. G. Lowe, "Distinctive image features from scale-invariant keypoints," International Journal of Computer Vision, Vol. 60, pp. 91-110, 2004.

    [4] D. G. Lowe, "Object recognition from local scale-invariant features, "IEEE International Conference on Computer Vision, pp. 1150-1157, 1999.

    [5] H. Bay, T. Tuytelaars, and L. V. Gool, "Surf: Speeded up robust features," Proceedings of European Conference on Computer Vision, Vol. 3951, pp. 404-417, 2006.

    [6] K. G. Derpanis, "Integral image-based representations, "technical report, Computer Science and Engineering York University, 2007.

    [7] X. Chen, H. Huang, and H. Zheng, C. Li, "Adaptive bandwidth mean shift object detection," IEEE Conference on Robotics, Automation and Mechatronics, pp. 210-215, 2008.

    [8] X. Chen, Y. Zhou, and X. Huang, C. Li, " Adaptive bandwidth mean shift object tracking," IEEE Conference on Robotics, Automation and Mechatronics, pp. 1011-1017 , 2008.

    [9] I. Dagher and K. Dahdah, "Adaptive bandwidth mode detection algorithm," IET Image Process, Vol. 5, Iss. 8, pp. 645-660, 2011.

    [10] D. Schreiber, B. Alefs, and M. Clabian, "Single camera lane detection and tracking," IEEE Conference on Intelligent Transportation Systems, pp. 13-16, 2005.

    [11] 任紹棟,"具立體視覺測距之移動機器人," 國立臺灣科技大學電機工程系碩士論文,2008。

    [12] C. Y. Lin and Y. P. Chiu, "The DSP based catcher robot system with stereo vision," IEEE/ASME International Conference on Advanced Intelligent Mechatronics, pp. 897-903, 2008.

    [13] W. Y. Yau and H. Wang, "Fast relative depth computation for an active stereo vision system," Real-Time Imaging, Vol. 5, pp. 189-202, 1999.

    [14] G. Bradski and A. Kaehler, Learning OpenCV: Computer Vision with the OpenCV Library. Sebastopol: O'Reilly Media, 2008.

    [15] Z. Zhang, "A flexible new technique for camera calibration," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 22, pp. 1330-1334, 2000.

    [16] M. Pal, A Course on Classical Mechanics. India: alpha science international ltd, pp.110-113, 2009.

    [17] G. Wei and S. Ma. , " Implicit and explicit camera calibration: theory and experiments, " IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 16, pp.469-480, 1994.

    [18] 繆紹綱編譯,數位影像處理。台北縣:普林斯頓國際有限公司,pp.385-446 2009。

    [19] 黃塏碩,"在非固定式鏡頭下之物件追蹤-基於顏色與紋理資訊的適應性粒子濾波器追蹤演算法,"國立臺灣科技大學電機工程系碩士論文,2009。

    [20] Y. Cheng, "Mean shift, mode seeking, and clustering," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 17, 1995.

    [21] T. Kailath, " The divergence and Bhattacharyya distance measures in signal selection," IEEE Transactions Communications Technology, Vol. 15, pp. 52-60, 1967.

    QR CODE