影像顯示科技蓬勃發展，其中的3D顯示技術在近年特別受到矚目。廠商推出了許多3D顯示商品，但仍無法受到消費者青睞。其中，3D影像內容不足，以及觀看3D影像的不舒適問題，是難以推廣的兩大因素。3D影像內容的製作需要耗費相當多的時間與成本，即時將2D影像轉換成3D影像的技術有可能改善3D內容不足的問題，但該技術是否能夠提供舒適、生動的立體視覺經驗是值得探討的問題。
本研究以市售的TriDef 3D影像轉換軟體為例，比較TriDef 3D在不同參數設定以及不同的顯示尺寸下對不同類型測試影像的視覺的感受。並利用3dsMax建立3D街景場景，比較TriDef 3D與雙視角立體顯示（Stereoscopic 3D, S3D）的視覺差異。為此，本研究設計了兩個人因視覺實驗：實驗一分別以23吋與65吋偏光眼鏡式立體顯示器呈現20張經TriDef 3D轉換的立體影像。受測者利用八分制李克特（Likert）量表評估不同深度與零平面位置參數下的多項視覺感受。實驗結果顯示深度設定（DP）與零平面設定（ZP）參數愈高，眼睛會越不舒適。相對而言，65吋螢幕的舒適度高於23吋螢幕。
實驗二以3dsMax建立一組10秒鐘的街景模擬影片，在65吋偏光眼鏡式立體顯示器上比較TriDef 3D與S3D的動態影像視覺感受。實驗結果顯示參數優化後的TriDef 3D在視覺舒適度上優於S3D。推測其原因在於參數優化後的TriDef 3D左右眼影像相對視差範圍較S3D小，雙眼影像比較容易融合，觀看場景中的遠近物體較不需大幅調變雙眼輻輳。

The development of image technologies is growing rapidly. 3D display technology has attracted extensive attention in recent years. Display manufacturers provided various 3D displays to the market. However, they are not widely accepted by the customers as lack of 3D contents and visual discomfort of viewing 3D displays. 3D contents normally are generated by costly 3D cameras or time-consuming 3D models using 3D software. Real-time 2D to 3D image conversion therefore is a favorite technology to copy with the shortage of 3D contents.
The aims of this study is to investigate the visual perception of viewing 3D images generated by a commercial “TriDef 3D” 2D-to-3D image conversion software with different parameter settings and different size of 3D displays. The study also compared the visual perception between the TriDef 3D and stereoscopic 3D (denoted as S3D) using a video clip with street scene generated by Autodesk 3dsMaxTM. Two psycho-visual experiments were conducted: Experiment 1 evaluated the visual perception on 20 image contents with depth (DP) and zero-plane (ZP) parameters using 8-grade Likert scale on a 23” and a 65” polarization-glasses type 3D display respectively. The results show the visual discomfort increased when the DP and ZP values are higher. In terms of the size of the 3D displays, 65” display is more comfortable than the 23” in watching 2D-to-3D image contents.
Experiment 2 compared the visual perception of a 10 second video clip between the parameter-optimal TriDef 3D and the S3D on the 65” 3D display. The results show the optimal TriDef 3D outperformed the S3D in terms of visual comfort. The possible reason is that the disparity ranges of the optimal TriDef 3D video are smaller than that of the S3D video. To fuse the left-right image in vision is easier and it doesn’t require dynamic convergence within foreground and background of the scene.

英文部分
1.S. Battiato, A. Carpa, S. Curti and M. Cascia, 3D Stereoscopic image pairs by depth-map generation, Proc. of 3DPVT, 124-131(2004).
2.C. C. Cheng, C.T. Li and L. G. Chen, Novel 2D-to-3D conversion system using edge information, IEEE Transactions on Consumer Electronics, 56(3), 1739-1745 (2010).
3.B. E. Goldstein, Sensation and Perception (6th ed.), Wadsworth (2002).
4.J. Y. Huang and P. L. Sun, Converting 2D photos to stereoscopic 3D images based on image segmentation, CVGIP 2009, 1315-1320, (2009).
5.Y. J. Jung, A. Baik, J. Kim and D. Park, Anovel 2D-to-3D conversion technique based on relative height depth cue, Proc. of SPIE,7237(2009).
6.J. Ko, M. Kim and C.Kim, 2D-To-3D stereoscopic conversion: depth-map estimation in a 2D single-view image, Proc. SPIE6696, A (2007).
7.M. C. Kao and T. C. Shen, A novel real-time 2D to 3D conversion technique Using Depth Based Rendering, IDW ’09, 2033-2034(2009).
8.C. T. Lin, S. C. Hsu, C. Y. Lin and J. F. Lee, A fully-automated 2D to 3D image conversion technique applied on stereo display, CGVIP2010, 1230-1231 (2010).
9.H. Murata, Y. Mori, S. Yamashita, A. Maenaka, S. Okada, K. Oyamada and S. Kishimoto, A real time 2D to 3D image conversion technique using computed image depth, SID Symposium, 29, 919-922 (1998).
10.C. S. Park and H.W. Park, A robust stereo disparity estimation using adaptive window search and dynamic programming search, Pattern Recognition, 34(12), 2573-2576 (2001).
11.P. L. Sun, T. Y. Chang and M. R. Luo, The Impacts of 3D Crosstalk on Visual Comfort, Proc. of 3DSA2012 (2012).
12.T. Shibata, J. Kim, D. M. Hoffman and M. S. Banks, The zone of comfort: Predicting visual discomfort with stereo displays, Journal of Vision, 11(8), 1–29 (2011).
13.W. Tam, A. Yee, J. Ferreira, S. Tariq and F. Speranza, Stereoscopic image rendering based on depth maps created from blur and edge information, Proc. of SPIE, 5664, 104-115 (2005).
14.W. J. Tam, C. Vázquez and F. Speranza, Three-dimensional TV: A novel method for generating surrogate depth maps using colour information, Proc. of SPIE, 7237, 1A (2009).
15.W. Tam and L.Zhang,3D-TV content generation:2D-To-3D Conversion, Proc. of IEEE ICME, 1869-1872 (2006).
16.O. Veksler, Stereo correspondence by dynamic programming on a Tree, Proc. of Computer Vision and Pattern Recognition, 384-390(2005).
17.Q. Wei, Converting 2D to 3D: A survey, Information and Communication Theory Group, Delft University of Technology, the Netherlands (2005).
18.S. Zhuo and T. Sim, On the recovery of depth from a single defocused image, Computer analysis of images and Patterns, Lecture Notes in Computer Science, 5702,889-897(2009).
中文部分
1.劉榮政(民90)，平面螢幕之立體影像設計，國立中央大學。
2.賴文能、陳韋志(民99)，淺談2D至3D視訊轉換技術,影像與識，16卷2期。
3.曾泰祥(民101)，大尺寸立體液晶顯示器的視覺舒適度範圍評估，國立台灣科技大學碩士論文。