Multimedia content distributed over existing communication networks are increasing in rapid pace. The traffic characterization of such information is greedy and requires special processing in order to be transmitted over exiting unreliable networks. Therefore, error resilience techniques have been proposed. These error resilience techniques affect the compression efficiency. The visual content varies rapidly during video transmission, which in turn influences the traffic characterization. MPEG-7 descriptors are used to represent the underlying media content. However, MPEG-7 descriptors have not been used in analyzing the performance of the video traffic. We propose a video transmission system that uses the motion intensity descriptors to ensure robust video transmission. A novel motion activity extraction technique is proposed by adopting a neural network approach. Our proposed extraction approach correlates well with human perception of the motion intensity of the video sequence. In order to emphasize the superiority of the proposed transmission system, we develop a selective packet dropping scheme that can be applied in case of network congestion.
|Original language||English (US)|
|Publication status||Published - Jan 24 2005|