三維立體影像轉換晶片

Stereo Image Based on 2D-to-3D Processor of Three-dimensional Display System

羅大哲(Ta-Che Lo)

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(2009年)

摘要

本篇論文所探討的是三維立體轉換晶片(2D-to-3D Processor)，此影像轉換晶片是一種基於多視點圖像所建立的，由多種角度所拍攝出來的影像，來產生三維立體影像，此晶片將來能夠整合於三維立體顯示器的電路中，可以讓使用者不需經過麻煩的立體輔助眼鏡，而是讓使用者直接用肉眼便可達到同樣且舒適的三維立體視覺效果。
針對立體影像顯示系統配合多視角(9-View)的三維立體顯示器，且不需經由深度影像資訊來達到具有深度視覺的立體影像，我們針對三維立體視訊的核心技術做研究，在三維視訊訊號源的取得上，我們提出了二維轉換三維的立體視訊轉換晶片，實現了與傳統二維視訊影像截然不同的新面貌。我們以現有的二維照片或是影片做為訊號源，利用精密的幾何計算，產生多視角的視訊來源，以特殊的子像素對映架構原理產生出三維立體影像，在配合上多視角的三維立體顯示器來使用播放，實現了三維立體視訊的轉換晶片。
本論文貢獻不僅可以有效的節省了傳統製作立體影像的後製時間，在未來二維顯示器只要附加這顆轉換晶片，並且輸入相關的視訊來源訊號，即可隨時觀看到真實且逼真的三維立體影像。

關鍵字

三維立體影像 ； 多視角 ； 立體顯示器

並列摘要

In this paper, a 2D-to-3D processor of three-dimensional system is discussed. This processor is a three-dimensional conversion with multi-view of autostereoscopic display. To generate three-dimensional stereo images, this processor integrates the optical elements needed for selective addressing of the two eyes in a remote display device, hence allowing free 3D viewing with the naked eye. In general, free-viewing 3D displays are more comfortable to the viewer.
This paper describes 2D-to-3D conversional processor which doesn’t need depth information to achieve depth of three-dimensional stereo image. We present an efficient mapping architecture using 3D slanted lenticular technology, and propose a simultaneous multiplexing method by using multiplex synthesis of the 9-view for autostereoscopic display. We provided an efficient structure for 2D-to-3D processor by using processing elements. A special mapping architecture of the sub-pixel structure produced the multi-view 3D image. The 2D-to-3D processor can be integrated with traditional display in the future.
We used TSMC 1P6M 0.18um technology to design our chip. Experimental results provide a comparison between the previous method and the proposed method. The simplify control scheme of regular data can be used to and reduce memory access. The advantages of these methods not only reduce the data access time but also achieve small area design.

並列關鍵字

Three-Dimensional Stereo Image ； Multi-view ； Autostereoscopic Display

參考文獻

[21] Ching-Yi Hsu, Yi-Pai Huang, Development and Researches of Real 3D Display Technologies, 2007

[2] Roberts, David E., ”History of Lenticular and Relate Autostereoscopic Method,” Leap Technologies, USA, 2003.

[17] K. N. Ogle, “Researches in Binocular Vision,” Hafner Publishing Co. Ltd, 1964.

[18] Qian Ning and Andersen A. Richard, ”A Physiological Model for Motion–Stereo Integration and a Unified Explanation of Pulfrich-like Phenomena” , Vision Res., Vol.37, No.12,pp. 1683–1698, 1997.

[19] “Laser Based 3D Volumetric Display System,” US Patent No. 5,854,613 (1998).