A rate-control scheme using fast frame-rate controlling and adaptive rate-quantization modeling for h.263

Abstract

Bit-rate control plays a critical role in thc video-coding. One of the main challenges of video-coding is to provide the guaranteed quality under highly compressed video. In our study, a structured R-Q base rate-control framework is proposed for low-delay video coding, which contains thrcc processing stages: the control of encoding frame-rate, bit-allocation for frame-level and the decision of quantization parameters for macroblock-level. First, thc proposed framework decides the encoding frame-rate per group of pictures (COPS) for better motion continuity. In the second stage, target-bit per frame is effectively estimated with the frame texture and buffer fullness. Last, based on an adaptive R-Q model modified by the Kalman filter for each cluster of MBs (COBs), a suitable.quantization parameter of each MB can be confirmed. In comparison with the TMNXrc module, our new framework can achieve higher objective and perceptual quality with lower computational complexity'. Index Terms H.263, Rate-Quantization Model, Kalman filter.