視訊編碼位元分配之研究

Abstract

隨著多媒體的普及化, 對於大量影音傳輸的需求亦與日劇增. 然而, 在有限的頻寬下, 我們必須尋找有效傳輸資訊的方式. 資料壓縮與資源分配是解決此困境的兩個主要議題. 在這裡, 我們主要研究如何將有限資源 做最佳化的分配. 有一些通訊的應用必須從同一個地點同時傳送多個影像資訊. 高畫值數位電視系統就是一個簡單的例子. 許多節目必須在經過一個有限頻寬的頻道同時傳送給用戶. 我們的目的, 就是要對這種多個節目的傳輸在一定的頻寬下做最佳化的分配. 我們考慮了多種不同組合的原始影像傳輸. 首先, 我們同時傳送同樣速率(例如:每秒十張)的數個原始影像. 這是最簡單的多影像資源分配問題,然而也是最重要的問題. 其結果可以證明我們的資源分配方法是否有效用.接下來, 我們考慮具有不同圖框率多個影像的傳送. 這是比較複雜的問題,但卻也是最能表現資源分配效用的例子. 最後, 我們考慮了對各個影像做屬於它自己的資源分配, 當然是在對各影像已施行資源分配之後所做的微調處理. 在與平均分配資源的結果比較後, 可以發現我們的資源分配確實在平均的影像品質有重大的改進.

Multimedia is more and more popular nowadays. It is muchrelated to the transmission of large amount of audio and video data. Limited bandwidth urges us to search the best way for transmitting data. Data compression and rate control are the two main issues. Here we make all efforts to distribute resource optimally and efficiently. Some communication applications involve the simultaneous transmission of multiple videos from one source location. For example, the HDTV system transmits many programs on a single channel with a limited bandwidth. We consider the optimal bandwidth allocation and coding of such co-located multiple videos for transmission over one constant bit rate channel. We outline an approach to optimal solutions under one type of distortion measure, namely MSE. However, we concern about many types of multiple video sources for transmission. First, we transmit all video sources with the same frame rate, which is the simplest case of multiple sources transmission but is the most important one that proves if we do the rate control in the proper way. Secondly, we transmit video sources with different frame rate. This is more complicated but is the case that we can get the most improvement if we do well. Finally, we consider local optimization in the single source under the rate control done in the previous way. The simulation results show a significant improvement in average MSE or PSNR compared to the coding which the resource is equally distributed.