我們的演算法對於主動式景深影像有很明顯的改善，尤其是在合成影像的結 果可看出改善前與改善後的差異。且我們的演算法大多簡單易懂，但仍有幾個可 供參考的未來工作，來做為改善此演算法的目標(1~2);對於環境的架設也有幾點 可供參考(4~6)。如下所示。
1. 在 C1(式(28))與 C2(式(33))可以由程式自動的決定最好的標準差數值，
讓影像的表現可以達到最大。
2. 在 C3 的參考方塊大小，可以自動的由程式決定最好的大小，讓整體的表 現達到最好。
3. 將此演算法從單一影像延伸至影片(Video)上。
4. 將此演算法用於不同種的景深感應器(如:SR4000、Xtion Pro)
5. 使用不同平台(如:OPENNI)產出的景深影像套用至我們的演算法上。
6. 針對主動式景深影像的缺點，開發出適合的合成演算法。
134
參考文獻
[1] A. Kubota , A. Smolic , M. Magnor , M. Tanimoto , T. Chen and C.
Zhang "Multiview imaging and 3DTV", IEEE Signal Process. Mag., vol.
24, pp.10 - 21, 2007.
[2] A. Smolic and D. M. Cutchen "3DAV exploration of video-based rendering technology in MPEG", IEEE Trans. Circuits Syst. Video Technol., Special Issue on Immersive Commun., vol. 14, pp.348 - 356, 2004.
[3] M.Tanimoto, “Overview of free viewpoint Television,” Signal Process. Image Commum, vol. 21, no. 6, pp. 454 - 461, July 2006.
[4] C.Fehn, “Depth-Image-Based Rendering (DIBR), compression and transmission for a new approach on 3D-TV,” Proc. SPIE Stereoscopic Displays and Virtual Reality Systems XI, pages 93 - 105, San Jose, CA, USA, January 2004.
[5] Microsoft,“XBOX360+Kinect,”http://www.xbox.com/zh-TW/Kinect/, 2012.
[6] Y.S. HO and E.K. Lee, “3DTV: Technical challenfes for realistic experiment,” in Interational Workshop on Advanced Image Technology, Jakarta, Indonesia, pp.
K3 1-6, Jan. 2011.
[7] K. Khoshelham and S.O. Elberink, “Accuracy and resolution of Kinect depth data for indoor mapping application,” Sensors, vol. 12, no. 2, pp. 1437-1454, 2012.
135
[8] J. Smisek, M. Jancosek, and T. Pajdla, “3d with Kinect,” in ICCV Workshop, Barcelona, Spain, pp. 1154-1160, 2011.
[9] Heresy, “OPENNI 環境/Kinect 安裝教學, ” http://kheresy.wordpress.com /2012/ 03/15/2012_openni_env_kinect_setup/, 2012.
[10] D. Herrera, J. Kannala, nad J. Heikkila, “Accurate and practical calibration of a depth and color camera pair,” in International Conference on Computer Analysis of Images and Patterns, Spain, pp. 437–445, 2011.
[11] Programming Guide: Getting Started with Kinect for Windows SDK Beta from Microsoft Research, Microsoft Research, July 2011.
[12] Marduk,“初探 Kinect 背後的 3D 測量技術,”http://blog.uns.org.tw/node/417, 2010.
[13] 曹祖聖,“Kinect for Windows 開發,”
http://msdn.microsoft.com/zh-tw/hh367958.aspx, 2012.
[14] Z. Zhang. “A flexible new technique for camera calibration,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 22, no. 11, pp. 1330-1334,
2000.
[15] A. Fusiello, E. Trucco, A. Verri , “A compact algorithm for rectification of stereo pairs,” Machine Vision and Applications.,vol. 12, pp. 16–22, 2000.
[16] F. Porikli, Inter-camera color calibration using cross-correlation model function,
136
in IEEE Int. Conf. on Image Processing, vol. 2, pp. II-133-6 ,2003.
[17] S. Lee and Y. Ho, “Real-time Stereo View Generation using Kinect Depth Camera,” in Proc. APIPA Annual Summit and Conference(APSIPA ASC), Xi’an, China, pp. RS3.12(1-4), Oct 2011.
[18] R. Hartley and A. Zisserman, Multiple View Geometry in Computer Vision, Second Edition, Cambridge University Press, March 2004.
[19] ISO/IEC JTC1/SC29/WG11, N11631, “Report on Experimental Framework for 3D Video Coding,” October 2010.
[20] PrimeSense, “ 3D Sensing Solution, ” http://www.primesense.com/en /technology /115-the-primesense-3d-sensing-solution, 2010.
[21] NUI API and Programming – Kinect for Windows SDK, “Image rectification between RGB and depth camera, ”http://social.msdn.microsoft.com/Forums/
en-US/kinectsdknuiapi/thread/40edbbc6-d0b1-4f14-b6a1-4a173c4dc588/, 2011
[22] H. Yu, G. Yang, Y. Zhou, Z. Jin, ”Improving Depth Estimation through Fusion of Stereo and TOF,” International Conference Multimedia Technology(ICMT), Beijing, China, pp. 26-28, 2011.
[23] C. Tomasi and R. Manduchi, “Bilateral filtering for gray and color images,”
International Conference on Compuer Vision(ICCV), Bombay, India, vol. 846,
pp. 839-846, 1998.
137
[24] Lai P., Tian D., and Lopez P., “Depth map processing with iterative joint multilateral filtering,” in Picture Coding Symposium (PCS), Nagoya, Japan, pp.
9-12, dec. 2010.
[25] M. Camplani, L. Salgado, “Efficient spatio temporal hole filling strategy for Kinect depth maps,” IS&T/SPIE Int. Conf. on 3D Image Processing (3DIP) and Applications , CA, USA, pp. 82900E 1-10, Jan. 2012.
[26] R. C. Gonzalez and R. E. Woods, Digital Image Processing, 3rd Edition Prentice Hall, 2008.
[27] A. Olofsson, “Modern Stereo Correspondence Algorithms: Investigation and evaluation,” Thesis from Dept. of Electrical Engineering, Linköping Univ., Linköping, Sweden, 2010.
[28] Microsoft Research, SkeletalViewer Walkthrough C++ and C#: Capturing Data with the NUI API, July 2011.
[29] K. Oh and Y. Ho, “Non-linear Bi-directional Prediction for Depth Coding,”
Advances in Multimedia Information Processing, Bangkok, Thailand, pp.
522-531, 2009.
[30] 劉楷哲、吳其霖、黃偉豪、陳信榮、李錕和羅豐祥，基於 3D 顯示器格式之 即時 3D 內容合成技術，影像與識別，第十六冊，第二期，第 76-95 頁，2010。
[31] Jean-Yves Bouguet, “Camera Calibration Toolbox for Matlab,”
138
http://www.vision.caltech.edu/bouguetj/calib_doc/, 2010.
139
自傳
邱義文，1988年1月3日出生於高雄市，高雄市人。2010年6月畢業於國立中興大學 電機工程學系。2010年7月於國立交通大學電子研究所攻讀碩士學位。承蒙杭學鳴 教授的指導，進入通訊電子與訊號處理實驗室(CommLab)。論文題目為「使用景 深感應器與RGB相機對用於影像合成的景深改善」