Research On Rate Control Algorithm For High Efficient Video Coding

With the rapid development of mobile Internet and multimedia technology,as well as the continuous improvement of living standard,people have put forward higher and higher requirements for video content and quality.As the rich content,high definition and ultra high definition video contain huge amounts of information,which make the amount of digital video data huge,digital video must be compressed before encoding or transmission,so as to reduce the demand for transmission bandwidth or storage space.In video communication,due to the limited transmission bandwidth of the communication network,the video encoder mostly adopts rate control technology to generate bitstreams that match transmission bandwidth and guarantee the video quality.A new generation of video coding standard HEVC adopts the rate control algorithm based on R-? model,which make the accuracy of rate control and video quality raised.However,the frame layer bit allocation does not consider the characteristics of video content and buffer state,which cause the frame layer bit allocation unreasonable,and the largest coding unit bit allocation fail to make full use of the correlation of the space time domain complexity,and the problems of model parameter updting process,which result in the largest coding unit layer bit allocation is not accurate.Aiming at the shortcomings of the rate control algorithm based on R-? model in the frame layer bit allocation and the largest coding unit layer bit allocation,two rate control optimization algorithms are proposed accordingly in the thesis.Aiming at the problem of the frame layer bit allocation unreasonable,a frame layer rate control optimization algorithm is proposed in the thesis.First of all,the image information entropy,the sum of absolute transformed difference and the frame layer fixed weight in the rate control algorithm based on the R-? model are weighted,which get a comprehensive factor that can measure the image complexity,and it is used as the new frame layer bit allocation weight.Secondly,calculate feedback bits in buffer according to the target buffer level and the remaining bits in buffer.Finally,the frame layer bit allocation is executed according to the obtained new frame layer bit allocation weight and the feedback bits in buffer.The realization and performance testing of the algorithm in this thesis are excuted on HM16.0.Experimental results show that the rate control error of the optimized algorithm is reduced by 2.008% averagely,and the peak signal to noiseratio is increased by 0.21 dB averagely,compared with the original rate control algorithm in the HEVC.Aiming at the problem of the largest coding unit bit allocation unreasonable,a largest coding unit rate control optimization algorithm is proposed in the thesis.Firstly,the gradient and the mean absolute difference of image are used to construct a new complexity.Secondly,an adjustment factor is defined by using the obtained new complexity and the gradient.Finally,the Newton method and video distortion are used to update the model parameters effectively.Experimental results show that the rate control error of the optimized algorithm is reduced by 0.062% averagely,and the peak signal to noise ratio is increased by 0.07 dB averagely,compared with the original rate control algorithm in the HEVC.