Rate Control Algorithm Based On HEVC

As the consumption level on video services increases,people are keen to pursue high resolution videos nowadays.However,the large amount of high resolution video data brings new challenges to video storage and transmission.At present,how to make the coding stream to meet the bandwidth requirement is a hot topic in video coding.Rate control becomes the key to solving this problem,through controlling the quantization parameter,the output rate is close to the bandwidth.In this paper,through the studies of intra frame rate control algorithms for the new generation of video coding standard HEVC,some algorithms and strategies are proposed.The contribution of this paper includes two parts:1)HEVC intra rate control algorithm does not consider the complexity of video content,rate distortion performance can be further enhanced.To solve the problem,this paper proposed the intra-frame rate control algorithm of HEVC based on the complexity analysis.Firstly,the gradient factor is used to represent the complexity of coding frame,the greater the gradient is,the higher the frame complexity gets.Bits are allocated reasonably for each frame according to the approximate linear relationship between image complexity and bits.Then the smoothing window is adaptively updated according to the difference between the actual allocated bits and the pre allocated bits to reduce the bit fluctuation.Finally,combined with the basic unit layer rate control algorithm adjusts the rate in time.Compared with Li's algorithm,for the same PSNR value,the proposed algorithm can save the bit by 2.3% on average,the maximal bit rate saving reach 2.9%.The objective video quality can increase 0.08 dB on average and 0.12 dB at most.The overall rate fluctuation is lower than Li's algorithm,and the bit error is further reduced.2)In order to effectively solve the problem that the rate control algorithm in HEVC does not take into account the video feature protection,a rate control algorithm based on feature protection is proposed.By acquiring the feature point mask,combining the three-layer bit allocation model of HEVC rate control algorithm.According to the distribution of feature points for each LCU assign different weights to improve the matching efficiency of LCU.Under the premise of not reducing the subjective and objective quality of video,the number of video feature points can be improved as much as possible.The experimental results show that compared with HM15.0 platform,within the acceptable range of time complexity,the proposed algorithm increases the matching rate by 1.27%.Hence more feature points can be protected and the performance of the algorithm is enhanced.