音视频压缩编码标准

简单介绍了音视频压缩编码标准的形成、内容及应用领域。着重介绍了MPEG-4的编码结构、编码算法。最后特别指出应遵守MPEGLA公司及其MPEG-4技术使用授权程序提案。     

一种内容自适应的硬判决量化算法

在视频编码器中,软判决量化可获得最优编码性能,但其中维特比算法会导致严重的系数间串行处理依赖.比较而言,基于死区的硬判决量化算法率失真性能有所损失,但不存在系数间依赖,且非常适合硬件并行流水处理.今采用离线统计的方法,分析余弦变换系数的分布情况,量化余数对量化的具体影响,构建了基于余弦变换系数分布参数,量化余数的自适应偏移模型.基于该模型提出一种内容自适应的硬判决量化算法.仿真结果表明,该算法在保证并行化处理的条件下,其编码性能相比于传统的硬判决量化有一定的提升,可接近软判决量化算法性能.     

MPEG-4音视频压缩编码标准

简单介绍了音视频压缩编码标准的形成、内容及应用领域.着重介绍了MPEG-4的编码结构、编码算法.最后特别指出应遵守MPEG LA公司及其MPEG-4技术使用授权程序提案.     

TCM模型的自适应硬判决量化算法

在视频编码中,DCT系数分布模型是率失真理论模型的基础,视频量化一般可分别为硬判决量化(HDQ)以及软判决量化(SDQ),SDQ算法能实现最优编码性能,但其中维特比算法会导致严重的系数间串行处理依赖.比较而言,基于死区(deadzone)的HDQ算法率失真性能略有损失,但是不考虑系数间的相关性.提出了一种基于分段逼近TCM模型(Transparent Composite Model)的自适应硬判决量化算法,采用更精确的DCT分布估计模型,估算不同频率分量DCT系数的分布参数.根据模型参数及DCT系数分布参数,优化构造自适应的死区偏移量模型.实验表明,相对于固定偏移量HDQ算法,其编码性能非常接近于SDQ算法.     

H.264码率控制算法改进

码率控制技术用来调整并控制视频编码器的输出比特率,以满足带宽要求和保持平稳的输出码流,并获取在质量和带宽利用上的最佳均衡。根据编码视频帧的复杂性及其编码比特来调解量化步长,达到降低最大码率的目的。结果表明,采用的该算法在不影响PSNR的情况下,码率极值得到一定程度的控制。     

低密度校验码量化译码研究

介绍了基于置信度传播算法的LDPC码和积译码算法，分析了译码中间变量的的密度进化特性，并对采用不同量化方案时LDPC码在AWGN信道下的译码性能进行了仿真研究。仿真结果表明相对连续译码．中间变量6bit均匀量化会带来约0．4dB左右的损失，而10bit非均匀量化性能明显得到改善。因此，合适的高阶量化译码可以获得接近连续译码的性能。     

一种新的自适应量化数字音频水印算法

文章中提出了一种新的自适应量化数字音频水印算法，该算法首先将视觉可辨的二值水印图像降维成一维水印序列，并对水印序列进行随机置乱与BCH纠错编码，再将原始数字音频信号划分成音频数据段，最后选择音频段进行快速傅立叶变换(FFT)，并依据人类听觉系统(HAS)模型自适应确定量化步长量化FFT系数嵌入水印信息。该算法提取水印信息时不需要原始数字音频信号。仿真结果表明：该自适应量化数字音频水印算法不仅具有较好的透明性，而且对诸如叠加噪声、有损压缩、低通滤波、重新采样等攻击均具有较好的鲁棒性。     

MPEG-4形状编码中的自适应二值运动估计算法

运动估计占据TMPEG-4形状编码90％以上的运算量，本文依据MPEG-4中形状信息的自身特性，提出了自适应二值运动估计算法。算法首先扫描参考帧，得出其有效搜索区域——边界掩模。在匹配运算时使用1bit的异或运算代替原有的加法运算。在搜索中根据二值Alpha块的运动特性自适应的采用不同的搜索模板进行搜索，对于静止点直接中止搜索。兼顾了搜索速度和精度。实验结果表明，在码字长度相仿的情况下，本算法的平均搜索点数仅相当于MPEG-4原有搜索算法的5％，有效提高了形状编码速度。     

一种基于TCQ的图像准无损压缩方法

图像准无损压缩一般包括预测、量化和熵编码三个部分。当前主流的编码方法CALIC和JPEG-LS主要通过提高预测的准确性,并对量化后的误差做有利于减小信息熵的变化来提高压缩质量,而对于预测误差的量化都只采用L∞准则约束下的最大步长均匀量化。本文通过引入性能更好的TCQ量化器,提出了一种新的准无损压缩方法。实验结果表明,新算法在低比特率时其率失真特性优于JPEG-LS。     

一种用于MPEG-4形状编码的快速运动估计算法

提出了一种新的形状快速运动估计算法,用于加快MPEG-4视频对象的形状编码过程。该算法充分利用了邻近形状块的运动向量之间的相关性、边界形状块的匹配特性。与MPEG-4验证模型中的形状运动估计算法以及现有的改进算法相比,该算法大大提高了形状运动估计的速度,而形状的编码效率并没有降低。     

拼凑块矢量量化方法

提出了拼凑块矢量量化编码方法,利用四步部分匹配预测技术获得一个拼凑块,作为当前图像块的预测被编码,只需一位便可以完成当前图像块的编码。在此编码方案当中,根据预先定义的匹配失真阈值,图像可以有选择地利用拼凑块、动态码本及码本库编码。四步部分匹配预测技术利用优化了每个图像编码块的像素空间连续性,改善编码性能,降低了误差传递效率。实验结果表明PBVQ在0.27bpp时得到了30.92dB的峰值信噪比,明显优于IFMFSVQ的29.82dB,CSMVQ的28.56dB,GSMVQ的28.64dB,很大程度地改善了重构图像质量,尤其在低比特率下。     

Classified vector quantization of LPC parameters

Abstract

To achieve high coding efficiency, modern speech coders adopt hybrid coding approaches, which utilize different coding mechanisms for various classified speech segments. With known voiced/unvoiced detection, in this paper, a classified LPC quantization (CLPQ) scheme is presented to effectively encode line spectral frequencies (LSF). The proposed CLPQ scheme improves the performance of the classified LSF vector quantizer, which adopts two LSF codebooks derived separately from voiced and unvoiced speech frames. With an objective spectral distortion measure, the CLPQ scheme successfully reduces the bit rate by about 1 bit/frame. Many classified LSF quantizers with different codebook structures and bit rates were evaluated. It would be helpful to design a classified LSF quantizer, which arrives at a compromise between distortion, bit rate and computational complexity.     

Variable‐rate hierarchical vector quantization with variance classification

Because the bit rate can be adjusted depending on the network traffic situation, the variable bit rate is more suitable than a fixed rate for transmitting images on the network. In this paper, a new variable rate hierarchical classified vector quantization (VHCVQ) scheme is proposed. This technique is suitable for progressive transmission. In our two‐stage VHCVQ scheme, we use the differences between the first‐stage input vectors and the first‐stage quantized vectors, called residual vectors, to code the image in the second stage. Usually the values between these differences are small and each is close to one another. Using the difference to code the image may reduce the codebook size and the bit rate as well. In our simulation, using these differences reduced the bit rate by about 0.075 bpp. © 2000 John Wiley & Sons, Inc. Int J Imaging Syst Technol 11, 138–143, 2000     

An effective codebook search algorithm for vector quantization

Abstract

A new scheme that aims to cut down on the computational cost of the vector quantization (VQ) encoding procedure is proposed in this paper. In this scheme, the correlation between the codewords in the codebook is exploited and three test conditions are designed to filter out the impossible codewords in the codebook. The design of test conditions is based on the concept of integral projection.

From the experimental results, it is shown that the new scheme outperforms all the other schemes proposed so far in speeding up the VQ encoding procedure. When the codebook of 1024 codewords is used in the proposed scheme, the execution time it consumes is less than 2 per cent of that needed by the full search algorithm. The average time reduction rate is approximately 97.7 per cent compared to the execution time for the full search algorithm. In other words, the proposed scheme indeed provides an effective approach to speed up the VQ encoding procedure.     

Effects of quantization in phase-shifting digital holography

We discuss quantization effects of hologram recording on the quality of reconstructed images in phase-shifting digital holography. We vary bit depths of phase-shifted holograms in both numerical simulation and experiments and then derived the complex amplitude, which is subjected to Fresnel transformation for the image reconstruction. The influence of bit-depth limitation in quantization has been demonstrated in a numerical simulation for spot-array patterns with linearly varying intensities and a continuous intensity object. The objects are provided with uniform and random phase modulation. In experiments, digital holograms are originally recorded at 8 bits and the bit depths are changed to deliver holograms at bit depths of 1 to 8 bits for the image reconstruction. The quality of the reconstructed images has been evaluated for the different quantization levels.     

莱斯信道下大规模MIMO系统混合模数转换器的随机向量量化方案

针对大规模MIMO系统用配置单精度模数转换器(ADC)来降低系统能耗会导致系统损失部分性能的问题,提出了一种莱斯信道下大规模MIMO系统的采用混合ADC的随机向量量化(RVQ)改进方案。该方案在频分双工模式下,首先在基站端采用高分辨率ADC和低分辨率ADC混合的接收方案处理信号,使接收的导频信号和有用信号具有较高的转换精度;系统对导频信号进行信道估计后,再对信道状态信息(CSI)进行RVQ处理,以此减小系统的反馈开销;最后运用最小均方误差(MMSE)信号检测算法减轻由量化误差引起的的用户间干扰,从而达到降低能耗并减小系统性能损失的目的。实验结果表明,这种改进的RVQ方案能在降低系统能耗的基础上有效减小系统容量损失,并使其和速率接近传统的高分辨率ADC接收方案。     

Bit rate reduction techniques for absolute moment block truncation coding

Abstract

Block Truncation Coding uses a two‐level moment preserving quantizer that adapts to local properties of the images. It has the features of low computation load and low memory requirement while its bit rate is only 2.0 bits per pixel. A more efficient algorithm, the absolute moment BTC (AMBTC) has been extensively used in the field of signal compression because of its simple computation and better MSE performance. We propose postprocessing methods to further reduce the entropy of two output data of AMBTC, including the bit map and two quantization data (a, b). A block of a 2×4 bit map is packaged into a byte‐oriented symbol. The entropy can be reduced from 0.965 bpp to 0.917 bpp on average for our test images. The two subimages of quantization data (a, b) are postprocessed by the Peano Scan. This postprocess can further reduce differential entropy about 0.4 bit for a 4×4 block. By applying arithmetic coding, the total bit reduction is about 0.3～0.4 bpp. The bit rate can reach 1.6～1.7 bpp with the same quality as traditional AMBTC.     

Embedded wavelet coder with multistage vector quantization

An embedded image coder based on wavelet transform coding and multistage vector quantization (VQ) is proposed in this research. We have examined several critical issues to make the proposed algorithm practically applicable. They include the complexity of embedded VQ, design of the successive vector quantizer, significance evaluation of a vector symbol, and integration of wavelet transform coding and vector quantization. It is shown in experiments that the new method achieves a superior rate-distortion performance. © 1997 John Wiley & Sons, Inc. Int J Imaging Syst Technol, 8, 444–449, 1997