摘要
H.264/AVC视频编码不能有效地处理场景切换,从而导致编码质量下降,码率控制偏差较大.针对这一问题,提出一种自适应变长GOP码率控制算法.首先用平均灰度差值检测场景切换,并提前结束当前GOP,将当前GOP未编码帧平均分配给后续两个GOP,然后修正码率分配和调整切换帧量化参数.仿真结果表明,改进的算法可以精确检测场景切换,提高编码质量,亮度PSNR平均增益达0.64d B,码率控制误差也更小,平均减少了2.71kbps的码率偏差,新算法对频繁发生场景切换的序列有较好的自适应性.
In order to effectively solve the problem of video quality droping and rate deviation is large when scene change is happen in H. 264/AVC video coding standard, an adaptive variable GOP length rate control algorithm is put forward. First detect scene change with the average difference of gray value and stop current GOP in the place of scene change, assigne unencoded frames in the current GOP average to the following two GOPs,then adjust bit allocation and quantization parameter. The simulation results show that ：the proposed algorithm is better than original rate control algrithm, such as detect scene change effectively, the quality of subsequent frames is improved with an average of 0.64dB of luminance PSNR,the error of rate control is also small,dropped an average of 2.71 kbps of rate deviation,new algorithm is well adaptability for sequences with frequent scene changes.
出处
《小型微型计算机系统》
CSCD
北大核心
2015年第1期188-192,共5页
Journal of Chinese Computer Systems
基金
江西省教育厅青年科学基金项目(GJJ11132)资助
关键词
场景切换
变长GOP
码率控制
平均灰度差值
量化参数
scene change
variable GOP length
rate control
average difference of gray value
quantization parameter
