一种对R-Lambda码率控制模型的改进算法
发布时间：2018-09-14 07:53
【摘要】：最新的视频编码标准HEVC的参考代码通常采用rate-Lambda(R-Lambda)模型进行码率控制,但R-Lambda模型为I帧分配的目标码率过多,导致后续视频帧因目标码率不足而重建质量严重下降.为此,对R-Lambda模型进行改进,根据图像的运动趋势和复杂度将I帧划分为运动区域、纹理复杂区域、纹理平坦区域.分配目标码率时,根据人眼视觉系统特性为视觉敏感区域分配更多码率,给非视觉敏感区域分配较少码率,最后利用R-Lambda模型计算相应的QP值.实验结果表明,所提算法在保证重建视频主观质量连续性的同时,提高了重建视频的整体质量.
[Abstract]:The reference code of the latest video coding standard HEVC usually uses the rate-Lambda (R-Lambda) model to control the rate, but the R-Lambda model allocates too many target bits to the I frame, which leads to the serious deterioration of the reconstruction quality of the subsequent video frames due to the lack of the target rate. Therefore, the R-Lambda model is improved and I frame is divided into moving region, texture complex region and texture flat region according to the motion trend and complexity of the image. When the target rate is assigned, more bits are assigned to the visual sensitive region according to the characteristics of the human visual system, and the less bit rate is assigned to the non-visual sensitive region. Finally, the corresponding QP value is calculated by using the R-Lambda model. Experimental results show that the proposed algorithm not only guarantees the continuity of subjective quality of reconstructed video, but also improves the overall quality of reconstructed video.
【作者单位】： 山东交通学院信息科学与电气工程学院;
【基金】：国家自然科学基金(No.61502277) 山东省自然科学基金(No.ZR2015FL017)资助
【分类号】：TN919.81
本文编号：2242052
[Abstract]:The reference code of the latest video coding standard HEVC usually uses the rate-Lambda (R-Lambda) model to control the rate, but the R-Lambda model allocates too many target bits to the I frame, which leads to the serious deterioration of the reconstruction quality of the subsequent video frames due to the lack of the target rate. Therefore, the R-Lambda model is improved and I frame is divided into moving region, texture complex region and texture flat region according to the motion trend and complexity of the image. When the target rate is assigned, more bits are assigned to the visual sensitive region according to the characteristics of the human visual system, and the less bit rate is assigned to the non-visual sensitive region. Finally, the corresponding QP value is calculated by using the R-Lambda model. Experimental results show that the proposed algorithm not only guarantees the continuity of subjective quality of reconstructed video, but also improves the overall quality of reconstructed video.
【作者单位】： 山东交通学院信息科学与电气工程学院;
【基金】：国家自然科学基金(No.61502277) 山东省自然科学基金(No.ZR2015FL017)资助
【分类号】：TN919.81
【相似文献】
相关期刊论文 前3条
1 潘伟;下一代核心光网络解决方案———Lambda Xtreme~(TM)Transport[J];电信科学;2002年11期
2 ;朗讯推出WaveStar Lambda全光路由器[J];世界电信;2000年02期
3 ;[J];;年期
相关重要报纸文章 前1条
1 ;DWDM的目标——Lambda交换[N];网络世界;2001年
，本文编号：2242052
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2242052.html
