在這篇論文中，我們設計和建立了一個雲端視訊處理平台，這個平台整合了視訊處理模組與雲端運算，程式開發者能夠很容易地產生一個分散式計算的雲端應用程式。這個提出的平台能夠達到下列目標：自動進行分散式的視訊處理，動態延伸雲端視訊處理功能，動態地啟動／停止雲端處理節點，根據目前的工作負載，以達到節能和符合服務延遲時間限制的目的。我們提出了個分析的佇列模型去決定最少需要多少處理節點去滿足計算的需求和服務層級協議，實驗結果證明了這個提出的平台的效能優於其它平台，對於總共需要的處理節點個數，能源的使用率以及處理節點的忙碌程度。

A cloud video processing platform is designed and built in this paper. This platform integrates video processing modules with cloud computing and programmers only need little efforts to create a distributed computing cloud application. The proposed platform can achieve these targets of automatically distributed video processing, dynamically extending cloud video processing functions as well as dynamically starting/stopping processing nodes based on the current workload for energy saving and satisfying the delay time requirement. An analysis queuing model is proposed to determine at least how many processing nodes are needed for meeting the computing requirement and satisfying the service level agreement. Experimental results demonstrate that the proposed platform outperforms other platforms in terms of total required processing nodes, energy utilization, and busy ratio.