9

Content-based Video Communications

9.1 Network-Adaptive Video Object Encoding

Network-adaptive video encoding is a robust video compression approach that designs and optimizes the source encoder by considering transmission factors such as error control, packetization, packet scheduling and retransmission, routing and error concealment. Error resilience is gained by selecting optimally the encoding mode that enables the decoded video to reach the minimum expected distortion for the available resources.

In this framework, we assume that the encoder knows the transmission channel characteristics such as the bit error rate (BER) or the probability of packet loss (in packet-based networks, the packet could be dropped or lost by bit error or excessive delay). This can be either specified in the initial negotiations, or calculated adaptively from messages exchanged by the transmission protocol. In a wireless channel, when transmission delay is not a major issue, the probability of packet loss can be calculated easily from the channel BER and the length of the packet. For ease of discussion, we assume that there is a lossy channel in this section and in section 9.2. In this way, packets are either received error-free or lost. The issues related to channels with BER will be addressed in section 9.3.

In order to evaluate received video quality, we assume that the transmitter knows the background VOP on which the transmitted video object will be composed at the receiver, which is possible ...

Get 4G Wireless Video Communications now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.