基于MPEG压缩域的快速、鲁棒的全局运动信息提取方法

全局运动信息在视频分析中起着重要的作用.本文根据MPEG编码特点,提出了一种从MPEG压缩域中快速有效地进行全局运动参数估计的算法.该算法充分利用了MPEG压缩码流中的信息,通过提取预测残差DC图像的运动背景区域,估计全局运动参数,从而保证了参数估计的准确性,有效地克服了已有文献中仅仅采用运动矢量进行全局运动估计的局限性.根据不同的MPEG测试序列的对比分析,结果表明,本算法可快速准确地对MPEG视频序列进行全局运动信息估计,同时具有很高的鲁棒性.     

MPEG—4中基于块的运动估计新算法

MPEG-4的应用和研究是当前的热点之一。运动估计是图像压缩编码的最重要环节之一，而基于块的运动估计方法是MPEG系列中使用最多的方法。MPEG-4组织推荐两种运动估计计算法：运动矢量场自适应搜索技术和可预测运动矢量场自适应搜索技术。然而，MVFAST算法搜索结果较差，PMVFAST需要一个强制的门限集合，且都是高于局部最优的方法。本提出了一种新的基于全局最优的遗传算法（GA：Genetic Algorithm)，实验表明，此算法运动搜索效果好，获得的图象质量高，有应用前景。     

运动估计算法的DSP实现研究

本文讲述了基于MPEG标准的运动估计的原理及实现,提出并用DSP系统实现了运动图像全搜索运动估计算法。实验结果表明,该系统能够快速有效地执行MPEG标准下序列图像的准确运动估计。     

MPEG-2到MPEG-4视频转编码中的快速运动矢量重估算法

本文讨论了MPEG-2到MPEG-4视频转编码问题，提出了两种快速运动矢量重估算法，分别基于空域和时域。我们的算法利用已有运动信息和DCT变换的直流系数进行运动估计，可明显减少运算量，提高编码速度，并获得了与全搜索算法接近的图像质量，可满足网络视频传输的实时要求。     

MPEG—2运动估计的实时实现

运动估计是ＭＰＥＧ图像编码中提高压缩率最有效的方法,也是运算量最大的部分,因而也是较难实时实现的。针对“高清晰度电视功能样机”信源编码的要求,就全局运动做主方法提出了相应的硬件设计,并且完成了其硬件实现。     

H．264运动估计算法优化研究

H．264是一种适应于低码率传输的新一代视频压缩标准,但其在获得高性能的同时,也导致了其复杂性很高。运动估计是运算量最大的一部分,对运动估计算法进行优化是提高编码效率的关键。基于此,从运动估计的基本原理入手,对几种典型的快速搜索算法进行了分析比较,在此基础上讨论了H．264标准中运动估计的特点和改进方法,并结合H．264标准的特点和快速搜索原理给出了H．264中运动估计快速算法的优化思路。     

一种用于块运动估计的匹配准则函数块特征匹配函数

本文提出了一种新的块运动估计匹配准则函数块特征匹配(BFM)函数,可以用于视频压缩的一些国际标准,如H.261,H.263,MPEG1,MPEG2,HDTV的编解码器中。在这些视频压缩国际标准中视频系统编码器的复杂性最主要取决于运动估计算法。实时的块匹配运动估计的VLSI实现需要考虑以下几个方面:在给定搜索域内运动搜索的复杂度;每次块匹配运算的匹配计算复杂度;每次块匹配运算需要从帧存读取到运动估计处理器的数据量大小;实时硬件实现的适用性.仿真表明BFM算法非常简单有效,可以大大降低相应的块匹配计算复杂度、匹配运算时数据传输时间.BFM函数便于并行实现,从而可以有效地缩短视频编码器的编码时间。本文还详细地给出了BFM函数与其它常用匹配准则函数的比较结果.     

MPEG-4自然视频编码技术

对MPEG－4自然视频编码的关键技术作了详细的分析和阐述。首先介绍MPEG－4频语法结构，并对视频编码的框架概要分析。接着分析了自然视频编码中涉及到的关键技术，包括：VOP的产生；二值和灰度级α平面的编码技术；运动估计和补偿方法；纹理编码；基于对象的时间分级和空间分级；MPEG－4提供的再同步和各种错误掩盖，刷新方法，精 编码技术和零树小波基的静止图像编码技术。指出MPEG－4和MPEG－1，MPEG－2等标准的异同，突出MPEG－4的三个主要特点。最后给出MPEG－4技术在网络视频传输中的应用，并给出测试和分析。     

基于MPEG压缩域的运动对象检测方法

为了从MPEG压缩码流中准确的检测和提取运动对象,本文提出了一种时空域的运动对象检测算法。算法主要利用了MPEG码流中的运动矢量信息,首先对运动矢量进行时域平均和向量中值滤波的预处理,减少运动估计秒准确带来的运动矢量与实际对象运动带来的检测误差。然后建立时域上关于搜索块与参考块之间运动矢量夹角的概率模型,对于帧间预测宏块通过聂曼一皮尔迅准则进行运动判决。同时,对于P、B帧内的编码宏块,提出判决算法区分运动的帧内宏块和重现背景。实验证明,本文算法可以获得较为理想的检测效果。     

基于MPEG的全局运动估计技术

提出了一种基于MPEG（moving Picture experts group)的全局运动估计技术;利用四参数模型来描述由摄像机运动带来的全局运动,再利用基于块匹配的算法来求解4个参数,进而作相应的运动补偿,由于MPEG视频流中的运动矢量蕴含了最优块匹配的信息,所以不需要对MPEG视频流完全解码,即可进行参数估计,实验证明该算法速度较快,并能通过运动补偿排除由摄像机运动造成的伪运动区域的干扰。     

Study on adaptive job assignment for multiprocessor implementationof MPEG2 video encoding

The high demand on computation performance by multimedia information processing such as digital video compression and decompression has made multiprocessor computation more and more popular. In this paper, we present our study on adaptive job assignment for multiprocessor implementation of a Motion Picture Expert Group 2 (MPEG2) video encoding algorithm. Data partitioning technique is used for job assignment to the processors in the multiprocessor environment to exploit the data structure adopted by the MPEG standard that divides a frame of a picture into macro blocks (MBs) which are processed independently during encoding. An adaptive data partitioning scheme is developed to cope with the inherent nonuniform spatial distribution of motion activities, such that the computation load distribution over processors is as uniform as possible, which helps improve the speedup of the whole multiprocessor system. Simulations with several video sequences have shown that, in comparison to its nonadaptive counterpart, the adaptive scheme can effectively improve the speedup of the multiprocessor system. In addition, the speedup scales well with the increase of the number of processors used in the computation     

No-reference analysis of decoded MPEG images for PSNR estimation and post-processing

We propose no-reference analysis and processing of DCT (Discrete Cosine Transform) coded images based on estimation of selected MPEG parameters from the decoded video. The goal is to assess MPEG video quality and perform post-processing without access to neither the original stream nor the code stream. Solutions are presented for MPEG-2 video. A method to estimate the quantization parameters of DCT coded images and MPEG I-frames at the macro-block level is presented. The results of this analysis is used for deblocking and deringing artifact reduction and no-reference PSNR estimation without code stream access. An adaptive deringing method using texture classification is presented. On the test set, the quantization parameters in MPEG-2 I-frames are estimated with an overall accuracy of 99.9% and the PSNR is estimated with an overall average error of 0.3 dB. The deringing and deblocking algorithms yield improvements of 0.3 dB on the MPEG-2 decoded test sequences.     

Fast and efficient motion estimation using diamond zonal-based algorithms

Motion estimation has always been an important part of video encoding systems because it can reduce temporal redundancy effectively and thus has significant impact on the bit rate and the output visual quality of the encoded sequence. Unfortunately, when using the brute-force full search algorithm, motion estimation consumes a very large portion of the encoding time. Previously, several algorithms have been proposed which try to reduce complexity, usually, with a significant loss in visual quality. Based on the diamond zonal search framework we introduced recently, we propose in this paper a novel algorithm called advanced diamond zonal search (ADZS), which was submitted to and well received by the Moving Pictures Experts Group (MPEG) standard committee for possible inclusion as an encoder optimization tool. ADZS was criticized in MPEG for using fixed thresholds, which may not be suitable for all video sequences. To address this issue, we further propose a threshold-adaptive version called threshold-adaptive advanced diamond zonal search (TAADZS). Simulation results verify the superior performance of ADZS and TAADZS over other fast algorithms and the robustness of TAADZS over ADZS.This project has been supported by the Research Grants Council of the Hong Kong Special Administrative Region, China (project no. HKUST6057/99E) and the Hong Kong Telecom Institute of Information Technology. Parts of this paper have been previously published in conference proceedings.     

Image/video spatial scalability in compressed domain

Image/video compression is widely used in various applications with the advent of the Joint Photographic Expert Group (JPEG), Motion Picture Expert Group (MPEG), and H.261 standards. Hence, most visual data are stored in the compressed format. Spatial scalable encoding of visual data has several applications, including browsing visual databases, querying multimedia databases, interactive multimedia communications, etc. We propose two novel techniques, namely, format compatible (FC) discrete cosine transform (DCT) and format modified (FM) DCT to implement image/video spatial scalability directly in the DCT compressed domain. The FC-DCT technique can be used to manipulate the standard bit streams, such as JPEG, MPEG, etc., while the FM-DCT technique can be employed in a variety of applications which require fast processing. In contrast to the traditional spatial-domain techniques, the compressed-domain techniques remove the unnecessary decompression and recompression procedures, thus, they have the advantages of reduced computational complexity and storage requirements. Simulation results demonstrate that the proposed DCT-domain techniques can achieve a comparable performance at a much lower computational complexity compared to the spatial-domain techniques     

H.264 motion estimation algorithm based on video sequences activity

Motion estimation is an important part of H.264/AVC encoding progress, with high computational complexity. Therefore, it is quite necessary to find a fast motion estimation algorithm for real-time applications. The algorithm proposed in this letter adjudges the macroblocks activity degree first; then classifies different video sequences, and applies different search strategies according to the result. Experiments show that this method obtains almost the same video quality with the Full Search （FS） algorithm but with reduced more than 95% computation cost.     

针对MPEG的去块效应技术的研究及其VLSI实现

现今随着数据压缩的广泛应用,以及MPEG视频编解码的普遍使用,块效应噪声也随之而来。为了消除块效应、提高视频质量,许多去块效应的算法[2-12]被人们所提出,然而大部分算法的运算量都很大,不适合应用在视频处理芯片上。针对这个问题,H.-S. Kong等人[13]提出了一种新的自适应后处理算法,该算法既能有效消除块效应又大幅度地降低了运算量,然而对图像边缘信息保护不够。本文在H.-S. Kong等人[13]的算法的基础上,提出了一种针对MPEG视频流的去方块后处理方法。该方法以方块边缘（水平或垂直）附近各两个4×4区域作为滤波处理的分类依据,通过对该两个区域的像素点的值进行计算,将块边缘分为平滑渐变区域、亮度阶跃区域以及复杂纹理区域,并根据这三种区域对图像边缘信息影响程度的大小,而决定采取强滤波、弱滤波或者不滤波等处理方式。这种方法一方面较好地提取了块效应噪声,一方面又保护了图像的边缘信号,而且计算量较小,满足视频处理芯片低复杂度的去噪要求,通过VLSI实现后有着很好的应用前景。      

MPEG coding for variable bit rate video transmission

For real-time transmission of broadcast-quality video on ATM-based B-ISDN, the intraframe to interframe ratio and the quantizer scale are two key parameters that can be used to control a video source in a network environment. Their impact on the traffic characteristics of the coder provides insights into the cell arrival process for an MPEG source. The authors obtain video source models for coders that utilize a standard algorithm that can be applied to a multitude of video services. The output stream of a video coder, which complies with the Motion Pictures Expert Group (MPEG) coding standard, is studied with an National Television Systems Committee (NTSC) quality video sequence as the input. Because the MPEG video coding algorithm has been proposed for a variety of applications, they also investigate the effect of changing the coding parameters on the statistics of interest     

A SEGMENTATION SCHEME FOR HEAD-SHOULDER VIDEO IN MPEG COMPRESSED DOMAIN

More attention has been paid to the study of video object segmentation in compressed domain these years, which has already led to some practical technology. In this paper, a scheme is put forward for segmentation of head-shoulder video in MPEG （Motion Picture Experts Group） compressed domain. The conception of DCT （Discrete Cosine Transform） feature plane is defined. In the suggested scheme, firstly, the face region is detected by clustering skin-tone DCT feature points in the DCT feature plane. Secondly, the region of head-shoulder is approximately regarded as combination of the head rectangle and shoulder rectangle, and head rectangle is confirmed by double template matching. Thirdly, Canny operator and morphological operation are applied to the region of head-shoulder in feature plane to get the object mask and the region of object mask is rectified by correlation of DCT blocks to get high-quality segmentation.