基于神经网络的视频编码量化参数选择算法

提出了一种基于神经网络的视频量化参数选择,利用已编码帧信息对当前帧的量化参数计算进行调整.神经网络模型选取当前帧的目标比特、当前帧的MAD和前一帧编码的QP和实际比特数作为输入,输出值经过反归一化及取整处理对视频进行量化.实验在H.264/AVC的标准参考软件JM14.2中实现,编码时使用该方法在保证码率精确度和平均PSNR值基本不变的情况下,减小了PSNR的波动.     

基于复杂度HEVC码率控制的算法优化

针对视频编码标准HEVC(high efficiency video coding)在编码过程中复杂度较高的帧层带宽相对不足而复杂度较低的帧层带宽相对富余之一的问题,为了降低计算复杂度,提高编码质量,本文首先通过研究复杂度在帧层编码过程中对码率分配的影响,提出了由SATD(sum of absolute transformed difference)构建复杂度模型,根据视频内容每帧的复杂程度,通过复杂度模型对帧层进行合理的比特分配。为了建立R-Q曲线,结合最大编码单元(LCU,largest coding unit)层的编码树特点,根据LCU残差信号的混合高斯分布特性进行量化参数(QP)的计算,从而使得输出码率与目标比特相等的同时尽可能提高视频质量。实验结果表明,所提出的算法比HEVC现行码率控制算法性能更加优良并且具有较强的鲁棒性,码率误差在0.01%以内,PSNR增益平均可达0.21dB。     

一种面向HEVC的并行码率控制算法

视频通信时通常带宽有限,为了能在规定的目标码率下获得尽可能高质量的解码图像,需要在视频编码时进行码率控制.目前针对HEVC的并行编码以及对应码率控制已成为研究热点,现有并行结构下的平均比特率控制算法受到帧间依赖性的约束,待编码帧无法及时获得与其并行编码帧的实际比特数,因此本文算法通过预测并行帧的实际比特数来进行码率控制,并在此基础提出了自适应调整帧层量化参数补偿值.仿真结果表明,相比已有算法,前者减少码率误差约为3.38％,后者可提高PSNR约为0.204 dB同时减少约0.3％的码率误差.     

不同量化比特数对超高清图像质量损伤的影响研究

通过挑选比较严格的超高清视频测试序列,采用x264压缩编码软件,计算量化比特数为10和8的各序列在不同码率下的PSNR值,以此来研究不同量化比特数对超高清图像质量损伤的影响。由实验数据得出,同一序列在同一码率下,量化比特数为10相比量化比特数为8的序列的PSNR高0.7 dB左右,也即量化比特数为10的序列相对量化比特数为8的序列码率节省5%~20%左右。     

H.264图像复杂度自适应的I帧码率控制算法

该文针对H.264帧内编码码率控制(RC)效果不佳的问题,提出了一种新颖的图像复杂度自适应I帧RC算法。首先用Sobel算子检测I帧亮度像素的梯度,建立44块的边缘方向直方图,得到每个44块最可能的帧内预测模式和相应重构块,最终获得与实际编码相近的残差图像。用残差的平均绝对值表达I帧编码复杂度,并提出了一种经验型码率-量化(R-Q)模型,同时考虑缓冲区状态和序列特性为I帧分配合适的目标比特,最后为每一个图像组(GOP)得到了合适的I帧量化参数(QP)。实验结果表明,该文算法能更准确地控制I帧输出码率,有效阻止缓冲区溢出和跳帧,得到更加平稳的视频质量,序列PSNR波动减少了60%以上。     

Wyner-Ziv视频编码中无反馈速率控制算法研究

为了避免在分布式视频编码系统中使用反馈信道,提出了一种基于Wyner-Ziv编码的无反馈速率控制算法。首先,利用目标码率和目标帧率进行GOP层码率分配;然后,根据原始图像的帧间相关性动态选择量化因子和量化矩阵来分配每个GOP内关键帧和Wyner-Ziv帧的比特数;接下来,利用系数带级的相关性计算相关噪声模型参数,并选择对应的LDPC校验矩阵,提出Wyner-Ziv帧的无反馈比特面速率控制算法。实验结果表明,在给定目标码率下,所提算法的编码码率误差小于0.57%,且与现有无反馈速率控制算法相比,解码恢复图像的PSNR(峰值信噪比)可以提高1dB。另外,该算法基本没有增加编码端复杂度,可用于实际分布式视频通信系统。     

一种改进的基于柯西模型的H.264码率控制方法

H.264/AVC编码中的码率控制是通过有效控制输出码流的码率来提高其压缩视频质量的重要技术。本文基于H.264/AVC中的JVT-H017码率控制方案提出了一种改进算法。新算法根据H.264中DCT系数的分布特征,将柯西分布引入到码率控制模块,用更精确的柯西率失真模型取代了原先的二次率失真模型。在此基础上,进一步引入了一种联合PSNR比率和MAD比率进行图像复杂度预测的方法,并依此来调整帧级比特的分配和量化参数,克服了在出现复杂运动或场景切换时,因视频序列相邻帧之间相关性降低而导致的MAD预测失准的情况。实验结果表明,与JVT-H017方案及文献[5]中的算法比较,新的算法不仅具有更精确的码率控制,而且明显改善了输出码率的平稳性及重建图像的PSNR。      

基于SSIM的HEVC帧内编码率失真优化

率失真优化(RDO)是视频编码压缩中的关键技术之 一。传统的率失真优 化技术使用误差平方和(SSE)或绝对差和(SAD)度量失真, 不能获得较好的视觉感知质量。本文针对新一代视频编码标准高效视频编码(H EVC)帧内编码RDO,提出了一种基于结构相似度(SSIM)度量失真的 RDO方法。首先,根据HEVC编码树形单元(CTU)结构设计了基于SSIM的失真计算方法; 然后,提出了一种基于量化器推导的码率-量化步长(R-Δ)模型和一种基于统计分析的失 真 -量化参数(DSSIM-QP)模型,用于求取拉格朗日乘数;最终,使用多QP优化方法求 取模 型参数。实验表明,相对于HEVC 传统的RDO,针对全I帧(即所有帧都为帧内编码)固定QP编码和多QP优化编 码,在相同的SSIM条 件下,码率分别平均降低8.4%和13.9%左右,同时编码复杂度分别平均增加约3%和2%。     

一种基于分层B帧架构的高效比特分配算法

随着分层B帧码率控制技术在SVC(可伸缩视频编码)中的广泛应用,视频编码的质量得到了很大提高,然而传统的H.264码率控制技术并未将B帧的比特分配考虑在内。提出一种基于分层B帧架构自适应调节B帧比特分配的码率控制算法,该算法研究了基于同一个图像组(GoP)中各时域层权重因子的自适应调节,为各时域层分配不同的目标比特;同时基于同一个时域层中各帧复杂度的不同,为各帧分配不同的目标比特。实验结果表明,与目前流行的分层B帧码率控制算法相比,该算法在比特控制误差并未降低的情况下,客观质量PSNR可提高0.1～0.4 dB。     

一种低复杂度的HEVC码率控制算法

针对HEVC视频编码中帧层比特分配和率失真模型参数自适应更新,提出了一种改进的码率控制算法.该算法引入牛顿法来自适应更新率失真模型的参数;同时提出一种帧层复杂度度量方法,利用图像自身的内容特性来调整帧层比特分配;最后在编码过程中选择合适的量化参数.实验结果表明:与参考算法比较,该算法能大大降低计算复杂度,同时保持良好的码率控制性能,对含有场景突变的视频改进效果尤为明显.     

Adaptive video coding control for real-time H.264/AVC encoder

In this paper, we present a new adaptive video coding control for real-time H.264/AVC encoding system. The main techniques include: (1) the initial quantization parameter (QP) decision scheme is based on Laplacian of Gaussian (LoG) operators; (2) the MB-level QP calculation is based on the spatio-temporal correlation, in which the computation is less than the quadratic model used by H.264/AVC; (3) the adaptive GOP structure is proposed, in which the I-frame is adaptively replaced by an enhancement P-frame to improve the coding efficiency; (4) the scene change is detected with the complexity of adjacent inter-frames and the appropriate QP is re-calculated for the scene-change frame. The proposed algorithm is not only to save the computational complexity but also to improve coding quality. Compared to the JM12.4 reference under various sequences testing, the proposed algorithm can decrease coding time by 64.5% and improve PSNR by 1.52 dB while keeping the same bit-rate.     

Window-based rate control for video quality optimization with a novel INTER-dependent rate-distortion model

Most model-based rate control schemes use independent rate-distortion (R–D) models at macroblock (MB) level to represent the relationship among bit rate, distortion and encoding complexity. However the correlations between frames (INTER-dependency) are not well considered for distortion, bit allocation and quantization parameter (QP) decision. In this paper, a novel INTER-dependent R–D model is proposed based on the theoretical analysis of the relationship between the predicted residual of one frame and the distortion of its reference frame. To achieve both bit rate accuracy and consistent video quality, a window-based rate control scheme with two sliding windows is introduced. One window is to group certain previously encoded frames and current frame to control the bit rate and buffer delay; the other is to group certain future encoding frames to optimize the fluctuation of video quality. Furthermore, the optimization of Lagrange multiplier is also discussed under the INTER-dependent situation. Experimental results demonstrate that the proposed window-based rate control scheme with INTER-dependent R–D model can achieve accurate target bit rate and improve PSNR performance, meanwhile the variation of PSNR is the smallest compared with other three benchmark algorithms. This one-pass rate control scheme is highly practical for the real-time video coding applications.     

Motion adaptive video coding scheme for time‐varying network

Adaptive video coding algorithms can encode a video stream dynamically on the basis of the amount of bandwidth available in a network. In this paper, a novel approach for adaptive video coding based on look‐up tables has been proposed. On the basis of the network conditions, the proposed codec estimates quantization parameter (QP) and also the spatial resolution for a group of pictures from the look‐up table. Then, QP for each frame is estimated on the basis of the motion content in each frame. More motion leads to more burden of bits while encoding the motion vectors. Also, quantization noise is less prominent in high‐motioned frames than frames possessing low motion information because of the temporal masking phenomenon of the human eyes. So, the main trick of the proposed scheme is to assign higher QP to the frames having higher motion than those having less motion. This method also reduces the requirement of excess bits for encoding of motion vectors having higher motion content. Additionally, QP is adjusted on the basis of the buffer availability in order to avoid bit loss error. Experimental results show the efficacy of the proposed codec. Copyright © 2013 John Wiley & Sons, Ltd.     

Human centered perceptual adaptation for video coding

Traditional visual saliency based video compression methods try to encode the image with higher quality in the region of saliency. However, the saliency feature changes according to persons, viewpoints, and distances. In this paper, we propose to apply a technique of human centered perceptual computation to improve video coding in the region of human centered perception. To detect the region of interest (ROI) of human body, upper body, frontal face, and profile face, we construct Harr and histogram of oriented gradients features based combo of detectors to analyze a video in the first frame (intra-frame). From the second frame (inter-frame) onward, the optical flow image is computed in the ROI area of the first frame. The optical flow in human centered ROI is then used for macroblock (MB) quantization adjustment in H.264/AVC. For each MB, the quantization parameter (QP) is optimized with density value of optical flow image. The QP optimization process is based on a MB mapping model, which can be calculated by an inverse of the inverse tangent function. The Lagrange multiplier in the rate distortion optimization is also adapted so that the MB distortion at human centered region is minimized. We apply our technique to the H.264 video encoder to improve coding visual quality. By evaluating our scheme with the H.264 reference software, our results show that the proposed algorithm can improve the visual quality of ROI by about 1.01 dB while preserving coding efficiency.     

New rate-distortion modeling and efficient rate control for H.264/AVC video coding

Rate control (RC) is crucial in controlling compression bit rates and qualities for networked video applications. In this paper, we propose a new rate-distortion (R-D) model and an efficient rate control scheme for H.264/AVC video coding, which elegantly resolve the inter-dependency problem between rate-distortion optimization and rate control by eliminating the need of coding complexity prediction for an inter-frame. The objective is to achieve accurate bit rate, obtain optimal video quality while reducing quality variations and simultaneously handling buffer fullness effectively. The proposed algorithm encapsulates a number of new features, including a coding complexity measure for intra-frames, a rate-distortion model, an accurate quantization parameter (QP) estimation for intra-frames, an incremental quantization parameter calculation method for inter-frames, a proportional+integral+derivative (PID) buffer controller, and an intelligent bit-allocation-balancing technique. Our experimental results demonstrate that the proposed scheme outperforms the JVT-G012 solution by providing accurate rate regulation, effectively reducing frame skipping, and finally improving coding quality by up to 1.80 dB.     

Rate control of MPEG video for consistent picture quality

We propose a rate control scheme using a rate-distortion (R-D) estimation model, which produces a consistent picture quality between consecutive frames. Our R-D estimation method offers a closed-form mathematical model that enables us to predict the bits and the distortion generated from a frame encoded at a given quantization parameter (QP) and vice versa. Its most attractive feature is its low computational complexity. Furthermore, it is accurate enough to be applied to practical video coding. In our simulation, the estimation errors for rate and distortion are less than 2.5% and 1.5%, respectively. Therefore, the proposed rate control scheme is appropriate for applications requiring low delay, low complexity, and the ability to control output bit-rate and quality accurately. Our scheme ensures that the video buffers do not underflow or overflow by satisfying the buffer constraint, and it also prevents quality difference between consecutive frames from exceeding a certain demanded level by adopting a distortion constraint. In addition, a consistent picture quality is maintained within a frame, and error propagation, caused by quality degradation of anchor frames, is reduced by differentiating the control procedure for anchor frames from that for nonanchor frames. Simulation results show that our control scheme achieves 0.52-1.84 dB peak signal-to-noise ratio (PSNR) gain over MPEG-2 Test Model 5 (TM5) rate control and maintains very consistent quality within a frame as well as between frames.     

A new rate control scheme for H.263 video transmission

In some image/video applications, the variable bit rate image/video bitstream will be transmitted over a constant bit rate transmission channel, in which a channel buffer is employed. In this study, a new rate control scheme for H.263 video transmission is proposed. Three proposed techniques include: preprocessing the INTRA coded macroblock (MB) incorporated the just-noticeable-distortion (JND) concept, constructing a bit estimation model in the frequency domain (instead of the bit estimation model in the spatial domain employed in Test Model Near-term version 11 (TMN11)), and adjusting quantization parameter (QP) for each MB by a Lagrangian optimization strategy.In the proposed approach, the target number of bits for a video frame is first obtained by using a simple rate control procedure for the frame layer. The proposed JND preprocessing is applied on all the INTRA coded MBs so that the number of coded bits for a scene changed frame will decrease, without any perceptual loss. Within the MB layer, instead of the bit estimation model in the spatial domain employed in TMN11, a bit estimation model in the frequency domain, directly depending on discrete cosine transform coefficients, is proposed. Then the Lagrange multiplier is used to determine the optimal QP for each MB. The resulting QP and number of coded bits of the current MB are fed backward to update the parameters of the bit estimation model.Based on the simulation results obtained in this study, the proposed approach can meet the target bit rate more accurately, keep a lower channel buffer fullness (delay), and have a larger average frame rate than TMN11, whereas the peak signal-to-noise ratio value and the processing time of the proposed approach are “approximately” as good as that of TMN11.     

A rate control scheme using Kalman filtering for H.263

This paper proposes a rate control framework for H.263 video coding called the low-delay Kalman filtering rate controller (LKFrc), whose main part is the adaptive Kalman filter. In LKFrc, a low-delay two-stage selector is proposed to select the representative frames of video. The first stage can rapidly pre-select the relatively significant frames. The second selection stage is embedded in the prediction model of Kalman filter. By the prediction model, LKFrc can distribute a proper bit budget to the pre-selected frame or definitely skip this frame. The measurement model of Kalman filter exploits a rate–distortion (R–D) model to meet the given bit budget by selecting an appropriate picture-level quantization parameter (QP). From the picture-level QP, the subsequent macroblock-level QPs are rapidly yielded by a rapid offset model using the relative macroblock activity. In comparison with the TMN8 model, the LKFrc framework can obtain better PSNRs, higher motion continuities and lower flickering effects with lower computation cost.     

基于多视点视频特征的3D-HEVC深度视频信息隐藏算法

为了更安全地传输秘密信息和保护3D视频,提 出基于多视点视频特征的3D-HEVC深度视频信息隐 藏算法。首先结合彩色视频的纹理特征,考虑深度视频不同区域对绘制视点的质量和编码效 率的影响,对 深度视频进行区域分割。然后针对不同的区域,采用不同的方式调制最大编码单元(LCU) 的QP值嵌入 秘密信息。最后,用修改后的QP值进行编码压缩,传输视频信息。实验仿真结果表明,相 比于原始HTM13.0, 本算法编码重建深度视频绘制视点质量的PSNR平均下降0.0015dB, 码率平均增加0.035%。本文算法 能较好地保证绘制视点的质量,对视频流的码率影响较小,且能实现秘密信息的盲提取。