即時視訊在無線網路上的傳輸，基於應用上的需要與狀況的變動性質，可調式音視訊編碼可提供可隨網路情況不同之變動性位元率的視訊資料，並在異質無線網路架構下，讓使用者選擇適合於各項無線行動裝置所支援的視訊服務與應用。而在無線網路上欲達成令人滿意的視訊服務品質，必須同時考量視訊處理在既定的系統負荷量下，所能提供的位元率與畫面品質，因此編碼端的系統複雜度與編碼時間，是另一個重要議題。基於可調式視訊編碼標準內所設計的編碼工具複雜度甚高，快速演算法以利其縮短編碼時間與降低複雜度的運用有其必要性。在位元率與解析度限制的情況下，視訊品質仍需獲得保證。應用超解析度技術（super resolution）於視訊壓縮編碼，將可得到更好的壓縮效益。本計畫預計在三年之中完成一可調式音視訊編碼系統，期能提供有效率的音視訊編碼機制，將服務品質提升。本計畫分為三大部分：第一部份專注於可調式視訊編碼器，以降低可調式視訊編碼器之計算複雜度為目標，進行提出演算法的各項研究步驟。第二部分將結合超解析技術，提升解碼端之視訊品質。第三部分研究可調式音訊編碼。研究完成之後，將會與其它子計劃整合以展示本系統的整體效果，相關成果並將發表於知名國際會議及期刊。 ; For real-time service over heterogeneous wireless networks, the scalable video and audio coding can provide variable bitrate bit streams with different frame rates, resolutions and media quality based on various channel bandwidth conditions. However, the computational complexity and power consumption of audio and video coding is critical under wireless environments. Thus, fast algorithms with low computational complexity are necessary to audio and video coding in this scenario. Alternatively, super resolution technique is supposed to be another scalable scheme that can be combined with video compression technique to improve quality. This project is expected to be finished in three years. It is aimed to develop a scalable video and audio coding system that provide efficient coding scheme to improve the quality of delivered video. This project consists of three major parts. The first part is to develop fast algorithms for scalable video coding. The second part will apply the super resolution techniques to video compression in order to improve the decoded video quality. Finally, the third part will work on the scalable audio coding. The outcomes of this project will be integrated with other projects to run on a heterogeneous wireless networks platform. ; 研究期間 9808 ~ 9907
