Video Encoder Optimization for Real - Time Communication

Abstract

Effective video encoder control ensures that the performance of a video delivery system can be optimized under system limitation. The focus of this work is on algorithm design for encoder rate and complexity control. On top of being useful tools during video encoding, these algorithms also aid in the study of system performance under different constraints.

If available computational resource does not allow the entire encoding process to be carried out in time, a complexity scalable technique that ensures a graceful degradation of coding performance will be a valuable tool. Such a video encoding scheme is also useful in embedded video encoders of ubiquitous mobile devices which operate under energy restrictions. We design a video encoding scheme which allows the rate-distortion process to be carried out in a complexity scalable fashion.

Effective rate and power control are especially crucial for the design of power constraint real-time wireless video encoders. Power invested in the encoder can potentially reduce rate and improve picture quality while transmitter power determines the capacity of the communication channel. Hence, it is not obvious how power should be allocated between the source encoder and the transmitter, or if controlling the allocation can bring significant performance improvement. A complexity scalable encoding scheme that allows the control of complexity coding performance trade-off enables the overall power optimization of a video encoding and delivery scheme. With an empirically derived complexity-rate-distortion model, both cross-layer resource allocation for individual user and resource allocation among multiple users can be studied.