基於DIBR轉換技術實現多視角立體影像顯示系統之晶片設計

Chip Design and Implementation of Multiview Generation System Based on DIBR

紀伯承(Po-Cheng Chi)

指導教授 ： 范育成

國立臺北科技大學/電資學院/電腦與通訊研究所/碩士(2011年)

摘要

有鑑於現今三維立體顯示技術應用正處於蓬勃發展的局面，各大廠也紛紛推出各式新穎的產品吸引消費者，然而不必透過輔助道具即可感受其擬真效果的技術，必然成為未來發展的首要目標，在此之前首先必須克服的可能包括觀賞舒適度及立體效果；然而改善立體影像的品質即為改善觀賞舒適度的重要關鍵之一。
本論文採用DIBR（Depth Image Based Rendering）轉換技術，產出多個視角之虛擬影像，藉由特定子像素對應架構合成出多視角影像。整個多視角立體影像顯示系統可分成四個階段作業：（1）將輸入的彩色影像與相對應的深度影像，利用DIBR轉換技術產出多個視角的虛擬影像；（2）當輸入深度影像不夠精確時，造成物件邊緣錯置，藉由加入物件邊緣修復演算法，有效提昇影像破洞填補效率；（3）藉由有效地填補影像破洞進而提昇多視角影像品質；（4）將多個視角的虛擬影像進行合成，產出能夠於相對應之立體顯示器上呈現立體擬真效果之多視角影像。
本篇論文主要的研究目標為：（1）希望藉由物件邊緣修復提昇多視角影像的品質，使觀賞者得到更優質的視覺感受；（2）希望藉由硬體架構的設計達到即時處理的效果，進而能夠應用於可攜式電子裝置上。

關鍵字

多視角影像 ； 虛擬多視角映射 ； 物件邊緣修復 ； 影像破洞填補

並列摘要

In view of the three-dimensional display technology is a flourishing situation nowadays, there are lots of various innovative products which are introduced to attract consumers by many famous relevant industries. If we can through the special means which brings three-dimensional reality to us by no other assistant tools, it will become the future development of primary objective. Before this, we must to overcome first some troubles about viewing comfort and three-dimensional effects, and stereoscopic image refinement is a key way to viewing comfort improvement.
We used DIBR (Depth Image Based Rendering) technology to generate virtual-views by a specific sub-pixel mapping to structure of multi-view respectively. The multi-view generation system can be divided into four sessions: (1) uses reference image and corresponding depth image generates the virtual-views through DIBR technology, (2) by adding the edge restoration algorithm to our proposed multi-view generation system, we can rise the efficiency of hole-filling when corresponding depth image is too rough, (3) there are many holes exposed after DIBR processing, we rise the quality of multi-view through proper hole-filling algorithm to fill them, (4) by the sub-pixel mapping structure, we can synthesize the multi-view from the virtual-views.
We studies the main objectives are: (1) hoping that through edge restoration algorithm to refine the multi-view, (2) hoping that through hardware design to implement multi-view generation system for real-time processing achievement, and thus can be implement in portable electronic device.

並列關鍵字

Multi-view ； DIBR ； Edge Restoration ； Hole Filling

參考文獻

[5] P. Kauff, N. Atzpadin, C. Fehn, M. Muller, O. Schreer, A. Smolic, and R. Tanger, “Depth Map Creation and Image-Based Rendering for Advanced 3DTV Services Providing Interoperability and Scalability,” Image Communication, vol. 22, no. 2, pp. 217-234, Feb. 2007.

[6] C. Fehn, “Depth-Image-Based Rendering (DIBR), Compression and Transmission for a New Approach on 3D-TV,” in Proc. of SPIE Stereoscopic Displays and Virtual Reality Systems XI, San Jose, California, USA, Jan. 19-22, 2004, pp. 93-104.

[9] L. Zhang, W. Tam, and D. Wang, “Stereoscopic Image Generation Based on Depth Images,” IEEE International Conference on Image Processing (ICIP ‘04), Singapore, vol. 5, Oct. 24-27, 2004, pp. 2993-2996.

[11] R-P. M. Berretty, F. J. Peters, and G. T. G. Volleberg, “Real Time Rendering for Multiview Autostereoscopic Displays,” in Proc. SPIE, California, USA, vol. 6055, pp. 208-219, Jan. 15, 2006.

[15] T. Fujii and M. Tanimoto, “Free-Viewpoint TV System Based on Ray-Space Representation,” in Proc. SPIE, vol. 4864, pp. 175-189, Jan. 2002.