Semantic-Based Video Transcoding Architectures for Quality of Service Applications in Mobile and Wireless Video Communication

Abstract

Delivering streaming video over wireless networks is an important component for most interactive multimedia applications running on personal wireless handset devices. Such personal devices have to be inexpensive, compact, and lightweight. Wireless channels have limited bandwidth and a high channel bit error rate and limited bandwidth. Delay variation of packets due to network congestion with the high bit error rate lessens the quality of video at the handheld device. Mobile access to multimedia content requires video transcoding functionality at the edge of the mobile network for interworking with heterogeneous networks and services. Under certain conditions, the bandwidth of a coded video stream needs to be drastically reduced. We present several efficient mechanisms for improving the quality of service (QoS) delivered to the client by deploying content-based transcoding schemes. The proposed approaches are performing the required transcoding based on the video content. Some approaches study the texture and temporal features. Other approaches perform object detection in order to determine the important objects to achieve semantic transcoding. The quality of the reconstructed images is remarkably similar to results that have been processed by the expensive and high performance transcoding approach. Exceptional performance is demonstrated in the experiment results. Extensive experiments have been conducted, and the results of various video clips with different bit rates and frame rates have been provided.