Evolution strategy based rate controlling for off-line video coding

Abstract

An intelligent rate control technique for H.264/AVC using the subjective quality of video is proposed. A subjective video quality metric is used with an evolution strategy algorithm, which is capable of identifying the best possible quantisation parameters for each frame/macroblock to encode the video sequence such that it would maximise the subjective quality of the entire video sequence subjected to the target bit rate. Simulation results show that up to 40% bit rate saving can be achieved with the proposed technique at the same video quality.