3-D MOD：網路立體視訊播放系統-子計畫三：支援可擴充性視角之立體視訊解碼器設計(I)

Abstract

多媒體視訊處理技術的蓬勃發展衍生出各式各樣有趣的應用，提升人們日常生活品質，也增 添了許多生活樂趣。從低複雜度靜態影像應用如數位相機、手機相機、數位相框等，發展到高複 雜度動態影像應用如DVD、數位電視、行動電視、數位攝影機、手持式多媒體播放器、網路視訊 與安全監控系統等，而效能需求也從低解析度手持式裝置應用等，發展到高解析度高畫質數位視 訊裝置應用等。而目前消費市場對於多媒體編解碼應用之需求也從現行的2-D視訊應用逐漸走向 3-D立體視訊應用，包含3-D立體數位相框、3-D立體照片、3-D立體電視、3-D立體遊戲等等。在 人們追求更高視覺享受與探索更新視訊處理技術之驅動下，視訊處理研究也從現行之2-D 視訊處 理技術發展到3-D立體視訊處理技術研究，3-D立體視訊處理帶來了更真實的視覺感受，也同時衍 生了更大的設計挑戰，包含更高的計算複雜度、更大的記憶體容量需求、更大的記憶體資料存取 頻寬與更大的操作功率消耗等。 為因應此技術發展趨勢，本計畫搭配總計畫所提出之全新的3-D立體視訊應用整合型計畫： 「3-D MOD：網路立體視訊播放系統」，進行子計畫三的研究，題目為「支援可擴充性視角之立 體視訊解碼器設計」，希望於此整合型計畫目標「3-D MOD：網路立體視訊播放系統」之架構 下(如圖11.1所示)，實現支援可擴充性視角之立體視訊解碼器設計，同時支援單一視角/多視角/任 意視角之立體視訊解碼功能，讓本計畫所提出之3-D立體視訊應用概念能夠逐步落實其技術可行 性，因此衍生各種創新之3-D立體視訊系統應用，以創造未來優質生活之藍圖。同時我們也希望 藉由此計畫，逐年完成數項具前瞻性之創新3-D立體視訊應用技術設計，以提昇國內學術界創新 技術之研發能量。 搭配總計畫之設計目標，本計畫在三年規劃中預計將可產出 (1) 單一視角立體視訊景深產生技術； (2) 單一視角立體視訊播放器設計； (3) 多視角立體視訊播放器設計； (4) 任意視角立體視訊播放器設計。

The technology evolution of multimedia video processing has induced various interesting applications, which has promoted the quality of people’s life and introduced lots of fun in everyday life. The technology development starts from the low complexity still image applications like digital cameras, mobile phone cameras, and digital photo frames, to the high complexity motion video applications like DVD players, digital TVs, mobile TVs, digital camcorders, portable multimedia players, network videos, and video surveillance. In addition, the performance requirement of the multimedia technology also evolves from low resolution portable video devices to high resolution high definition TV devices. Moreover, the market requirement on multimedia products also gradually shifts from the 2-D video products to 3-D video applications, including 3-D digital photo frames, 3-D photos, 3-D TVs, 3-D video games, and so on. Under the driving forces of pursuing better visual quality and exploring better video processing technology, the research on multimedia video also shifts from 2-D video technology to 3-D video technology. The 3-D video technology will definitely bring better visual quality, but it also induces much higher design challenges like much higher computational complexity, much larger memory requirement, much larger memory access bandwidth, and much larger power consumption. Following this technology trend, this project is the sub-project 3 of the main project entitled “3-D MOD: network 3-D video playing system” for bringing a brand new video applications in the coming future. The title of this sub-project 3 is “Multi-view Video Decoder Design Supporting View Scalability”. Under the design concept of the main project, as shown in Fig. 11.1, this sub-project will develop a view scalable multi-view video decoder that can support single-view/multi-view/free-view point video decoding capability. The goal of this sub-project 3 is to enable the possibility of innovative 3-D video processing techniques into mass production so as to induce lots of brand new video coding applications in the near future. Moreover, we will develop some advanced video coding technology in this project to enhance the research capability of Taiwan in this area. To support the main project, this sub-project 3 will deliver the following key technology in three years: (1) Single view video depth generation technology; (2) Single view video player design; (3) Multi-view video player design; (4) View scalable multi-view video player design.