基于非局部低秩和自适应量化约束先验的HEVC后处理算法

经高效视频编解码标准HEVC压缩后的视频在高压缩比、低码率的情况下存在明显的压缩效应。针对该问题,提出了一种基于非局部低秩(Non-local Low-rank, NLLR)和自适应量化约束(Adaptive Quantization Constraint, AQC)先验的HEVC后处理算法。该算法首先构造在最大后验概率框架下的优化问题,然后利用解码后的压缩视频和量化参数QP获取非局部低秩和自适应量化约束先验信息,最后利用split-Bregman迭代算法来解决所提的优化问题,从而有效去除压缩效应,提升重建视频质量。其中,非局部低秩先验通过构建基于相似块聚类的非局部低秩模型来获得;自适应量化约束先验通过联合不同量化参数QP下的约束特性与视频的DCT域块活动性来获得。实验结果表明,在同等码率的情况下,与HEVC标准相比,所提算法在帧内编码模式下可以达到平均0.259 7 dB的PSNR提升,在帧间编码模式下可以达到平均0.282 8 dB的PSNR提升。     

基于纹理相似性的高效视频编码帧间预测优化算法

为了提高高效率视频编码(HEVC)的编码效率,该文对帧间预测算法进行优化,提出一种基于纹理相似性的快速深度判决算法。随着视频分辨率的提高,视频序列中的空域冗余也随之增加,HEVC通过增加编码块尺寸来提高预测效率,代价就是编码复杂度的急剧增加。通过对视频序列分析可知,图像中的平滑区域与细节区域在相邻帧中存在很大的相关性。该文利用相邻已编码帧的相邻编码块的深度信息,来快速判决当前待编码块的深度信息。对于平滑区域,快速判决待编码最大编码单元的最大深度,以减少小块编码单元的模式判决;对于细节区域,快速判决待编码最大编码单元的最小深度,以减少大块编码单元的模式判决。实验结果表明,与原始的HEVC算法相比,该文所提算法编码比特率平均增加约0.13%以内,PSNR的平均降幅为0.09 dB,算法运行时间平均减少了约50%。     

基于感兴趣区域的高性能视频编码帧内预测优化算法

针对高性能视频编码(HEVC)帧内预测编码算法复杂度较高的问题,该文提出一种基于感兴趣区域的高性能视频编码帧内预测优化算法。首先,根据图像显著性划分当前帧的感兴趣区域(ROI)和非感兴趣区域(NROI);然后,对ROI基于空域相关性采用提出的快速编码单元(CU)划分算法决定当前编码单元的最终划分深度,跳过不必要的CU划分过程;最后,基于ROI采用提出的预测单元(PU)模式快速选择算法计算当前PU的能量和方向,根据能量和方向确定当前PU的预测模式,减少率失真代价的相关计算,达到降低编码复杂度和节省编码时间的目的。实验结果表明,在峰值信噪比(PSNR)损失仅为0.0390 dB的情况下,所提算法可以平均降低47.37%的编码时间。     

Pure intra mode decision in HEVC using optimized firefly algorithm

High Efficiency Video Coding (HEVC) is the latest encoder that increased the intra modes from 9 to 35 to efficiently handle the contents of the video. The HEVC’s test model (HM) selects the optimal intra mode using the brute force method which increases the complexity of HEVC. This work, firstly, investigates the feasibility of firefly algorithm (FFA) due to its exploration and exploitation characteristics to expedite the intra mode decision in HEVC. Secondly, a novel objective function is formulated for FFA to efficiently compute the brightness for intra modes in FFA. Thirdly, the parameters of FFA are made dynamic to adjust according to the contents of video sequences. Simulation results demonstrate that the nature inspired algorithm, FFA, pays off by saving a minimum of 27% of the total coding time on average and doesn’t sacrifice quality by limiting Bjontegaard delta bit rate (BD-BR) increase to only 0.98% on average.     

一种HEVC帧内预测算法的动态自重构实现方法

高效视频编码HEVC中帧内预测算法在专用硬件上的实现无法满足在高清和移动视频等多种应用场景间灵活切换的需求,导致编码性能差,硬件资源利用率不高。针对这一问题,提出一种新的帧内预测算法在可重构阵列处理器上的实现方法。该方法基于状态监测机制监测处理单元的执行状态,监测到空闲状态的处理单元则下发新的执行任务,根据处理单元的执行状态实现不同映射方案间的灵活切换,达到算法执行过程的动态自重构。实验结果表明,与帧内预测算法在专用处理器上的实现相比,本文方法在提高灵活性的同时,硬件资源使用减少了33.6%,算法执行的时钟周期数减少了16.2%。不同测试序列经过整个I帧环路测试的结果,与HM16.7官方软件的测试结果相比,平均图像质量有所提高。     

Adaptive quantization parameter selection for low-delay HEVC via temporal propagation length estimation

Rate Distortion Optimization (RDO) is employed in High-Efficiency Video Coding (HEVC) to enhance its coding efficiency. Owing to its coding complexity, RDO is traditionally performed within each frame with fixed quantization parameters (QPs), without fully considering the coding dependencies between the current frame and future frames within a temporal propagation chain. To improve the coding efficiency of HEVC, it is desirable to perform a global RDO among consecutive frames while maintaining a similar coding complexity. To address this problem, in this paper, temporal dependencies are first measured via a model for the energy of prediction residuals that enables the formulation of the global RDO in low-delay (LD) HEVC. Second, the notion of propagation length is introduced, which is defined as the impact length of the current frame on future frames. This length is estimated via offline experiments and used to propose two novel methods to predict the impact of the coding distortion of the current frame on future frames from previous frames of similar coding properties. Third, we apply these two methods to adaptively determine the Lagrangian multiplier and its corresponding QP for each frame in the LD configuration of HEVC. Experimental results show that, in comparison to the default LD HEVC, the first method can achieve, on average, BD-rate savings of 5.0% and 4.9% in low-delay-P (LDP) and low-delay-B (LDB) configurations, respectively, and the second can achieve, on average, BD-rate savings of 4.9% and 4.9% in the LDP and LDB configurations, respectively, all with insignificant increases in encoding time.     

结合内容特性与纹理类型的HEVC-SCC帧内预测快速算法

本文提出了一种结合内容特性与纹理类型的HEVC-SCC帧内预测快速算法。利用自然内容和屏幕内容视频DCT变换后系数能量分布不同的特点,结合当前预测单元（Prediction Unit, PU）梯度信息,将编码树单元（Coding Tree Unit, CTU）分成自然内容CTU,简单屏幕内容CTU和复杂屏幕内容CTU。对于自然内容CTU,选择35种传统帧内模式作为候选模式,跳过帧内块复制（Intra Block Copy, IBC）和调色板（Palette mode, PLT）模式;对于简单屏幕内容CTU,选择DC,PLANAR,水平和垂直模式作为候选模式,跳过IBC和PLT模式;对于复杂屏幕内容CTU,选择IBC和PLT模式,跳过其他候选模式。实验结果表明,在全I帧条件下,该算法相较于SCM-8.3可以节省38.55%的编码时间,大幅度降低了编码复杂度的同时只增加了1.82%的码率。      

基于HEVC的CABAC硬件解码器设计

首先分析了高效视频编码(High Efficiency Video Coding,HEVC)编码标准,对HEVC的新特性进行总结,并给出HEVC解码器整体架构。其次提出基于熵解码部分优化的硬件解码器架构,并进行仿真验证。最终通过与HM12.0软件评测结果的对比,证明所设计的上下文自适应二进制算数编码(Context Adaptive Binary Arithmatic Coding,CABAC)硬件解码器性能能够满足解码要求。     

基于改进的Canny算子的HEVC帧内模式选择

针对HEVC中帧内预测模式选择算法复杂度问题,提出了一种基于改进的Canny算子的HEVC帧内模式选择算法. 首先,通过改进的Canny边缘检测算子提前分析预测块的方向,将此方向相应的角度预测模式加入到候选模式集中;其次,利用图像空间相关性,将相邻预测单元的最优帧内预测模式添加到候选模式集中,最后,通过RDO代价进行帧内模式选择. 实验结果表明,与粗选择模式相比,该算法将候选模式数目从35种减少到不超过9种,算法编码时间平均减少32%,而BD-Rate损失仅为0.096%,并且编码后的主观图像质量基本没有变化.