Improved Rate Control Algorithms For Encoding Standard H.264

H.264 video coding standard is provided formally by ITU-U/ ISO. The framework of H.264 is similar with all video encoding standards. And many new technologies, such as spatial prediction intra encoding, adaptive block size motion compensation, 4×4 integer transformation, multiple reference pictures and content adaptive binary arithmetic coding (CABAC), are used in it. The signal/noise ration obviously raised in same bit rate in this standard. So it obtains much attention from academic and industrial fields.And there are difficulties of encoding because of the new technology. For example, rate control of H.264 becomes difficult because of the chicken and egg dilemma which brought by the quantization parameter used for rate distortion which determines the best coding mode and the rate control. So the new rate control algorithms are designed to deal with this problem.And there are defaults of new algorithms for H.264 rate control faced to the move fast sequences, such as average allocating bit rate and inaccurate MAD prediction. So this dissertation proposes an improved rate control algorithm for H.264 based on the detection of science change and movement. The computing of rate control model exploited by H.264 is complex. And the imprecise rate control is brought by it. Therefore a nonzero coefficient based on a map is proposed to deal with these defaults.Firstly, an improved rate control algorithm is proposed based on the detection scene change and high motion. The improved rate control algorithm optimizes linear MAD prediction model in encoding process when the scene change or high motion pictures are detected. And then, this optimized MAD will be used to computing an optimized quantization. And bit rate allocating is according to the comparability of frames in a same complication.Lastly, a nonzero rate control based on a map is discussed. The relationship between nonzero and bit rate is studied to optimize rate to control bit rate, combined with a quantization and rate relationship map.The simulation shows that the PSNR of improved rate control algorithm base motion detection is obviously improved for the pictures with high motion by compared to JVT-G012. And the PSNR of nonzero rate control based on a map is obviously increased when the bit rate is high by compared to JVT-G012. Furthermore, this algorithm conquered the inaccurate rate controlling of JVT-G012 when the bit rate is low. So this algorithm is stronger than JVT-G012 when the bit rate is low.