Motion assisted frame complexity estimation based video rate control for smooth quality

Rate control is an essential technique for video codec designing. For the purpose of smoothing video output quality, more bits should be allocated to the frame with high coding complexity. A novel coding complexity prediction algorithm based on motion extrapolation is proposed, and combined with coding buffer occupancy to determine the frame-layer bit allocation, which could smooth the video output. Simulation results show that compared with the existing frame-layer video rate control schemes, the proposed one could significantly reduce the video quality jitter up to 20% with pretty little additive computing cost, thus reach the aiming rate with more accuracy. The proposed algorithm could be combined with the rate control algorithm for basic unit to provide accurate video rate control.