计算机科学与技术 英文 外文 文献翻译 JPEG2000图像编码系统理论与应用.doc

（文献翻译） 第 14 页 JPEG2000图像编码系统理论与应用摘 要JPEG2000是新一代静止图像压缩标准，它提供了一个新的框架和一个综合性的工具箱，以便于更好地解决人们对数据压缩的要求。它提供了广泛的功能，如无损和有损编码，嵌入式损耗到无损编码，分辨率和质量的可扩展性，高质量，高压缩效率，错误恢复能力和感兴趣区域（ROI）的编码。结果表明，JPEG2000标准确实优于建立图像压缩标准。总体而言，JPEG2000的标准提供了丰富的功能，一个非常有效的方式和范围内统一的算法。由于技术的迅猛发展，它的价格有时带来不少压力，但这不应视为弊端。一. 引言 国际标准JPEG2000是代表图像压缩技术在图像编码系统的优化，不仅提高效率，而且在网络和移动环境的可扩展性和互操作性的发展。数字图像已经成为互联网不可分割的一部分，JPEG2000也是一个能提供设计和网络图像应用的强大新工具。随着多媒体和互联网应用对图像编码技术的需求不断增长和发展，自1997年3月起，需要发展一个新的静态图像压缩标准。JPEG2000是JPEG工作组制定的最新的静止图像压缩编码的国际标准，标准号为ISO/IEC 15444（ITU-T T.800），并于2000年陆续公布。第1部分是标准的核心编码系统，是由有限数字的编码算法组成，以便提供最大限度的交流。第2部分是编码扩展，规定了核心编码不具备的功能扩展。显然，第2部分将无法被第1部分的解码器译码。作为一个例子，第2部分包括DC电平位移，多种分量变化，网格编码量化，用户定义的小波，任意小波分解，一般基本压缩比例的感兴趣区域编码和良好的抗误码性方案（图1）。它实际上是一个对各种专门技术所应用的工具箱。第3部分所定义的运动JPEG2000（MJ2或MJP2），是根据第1部分JPEG2000的。MJ2将被用于许多不同的领域，例如一个单一的编解码器为静止图像和运动序列（一个共同特点，数码相机），或在非常高品质的动态图像需要（如医疗图像和电影制作），或在视频中容易出错的环境（如无线和互联网）等所应用。JPEG2000的压缩图像序列，同步音频和数据都可以被存储为MJ2文件格式。运动JPEG2000标准也将目标瞄准了与JPEG2000的文件格式（JP2）和MPEG-4文件格式（MP4）的互操作性。一个新的专案小组成立于2002年10月以促进JPEG2000标准在医疗界上的应用（运动JPEG2000的医学影像）。第4部分标准定义了一致性测试。第5部分定义了参考软件（高质量的免费软件）。提供了实现标准的两个软件，也就是JPEG2000在Java软件和JasPer软件在C语言上的使用。Kakadu软件也是可以的。第六部分规定了混合图像文件格式。第7部分，原打算规定一个具有支持最低限度功能的（第一部分）技术报告，但已被撤销。第8部分涉及安全问题，如认证、数据完整性、保护版权和知识产权、隐私权和有条件的访问。第9部分定义了互动工具，应用程序接口和协议为JPEG2000图像以一个客户端/服务器环境的方式传输。第10部分，限定了JPEG2000的算法（对三维图像压缩和浮点数据）。第11部分，定义了一个压缩的图像传输在无线环境中的文件格式，如移动电话，无线局域网和无线电台。第12部分，定义了MPEG-4和运动JPEG2000的共同文件格式。第13部分，确定一个规范性的入门级JPEG2000编码器，以提供一个或多个可选的完整编码路径，从而能使用各种功能中所规定的JPEG2000标准。现在JPEG2000是比JPEG格式更为复杂，它依赖于离散小波变换（DWT）和编码在大块小波系数上，所以比JPEG算法成本更高。嵌入式熵编码算法是远离琐碎执行的代码流解析规则涉及大量复杂的概念。JPEG2000最苛刻的部分是DWT和熵编码。许多架构无论是硬件和软件，迄今为止都是提出了为减少计算和内存要求。严格的架构JPE2000以编码效率预计，这是远远高于其他编码方法或标准，同时支持无损和有损。 二 JPEG2000压缩架构 图1显示，图像如何通过不同阶段的JPEG2000编码过程而转变。值得一提的是，与许多编码方法不同，JPEG2000压缩可以是有损也可以是无损。JPEG2000标准还允许图像分片。“分片”指的是把原始图像（源）分割成相互不重叠的、压缩独立的矩形块（片），好像它们是完全不同的图像似的。图像片允许任意大小的分片，但包括整个图像（即没有图像片）。它降低了储存空间的要求，是有效提取某部分图像的方法之一。A.小波变换 通过对图像片进行离散小波变换，得到小波系数图像。分解级数是通过两个方向的空间频率关系到下一个分解级数。第1部分只支持标准二进制分解，因为这能体现具有最佳空间分辨率的原始图像。DWT可以是不可逆也可以是可逆的。不可逆的变换在JPEG2000第1部分是通过 9-tap/7-tap滤波器进行，可逆转变是通过5-tap/3-tap滤波器进行。B.量化 在编码器，分过等级的量化操作图的某一值系数作为量化指标，然后对编码的一部分进行压缩比特流。在解码器，量化指标是解码并转换成相应的量化值。量化操作是有损的，除非量化步长是1，并且小波系数都是整数，作为可恢复整数5 / 3拍小波过滤器。子带b的小波变换系数 量化为量化系数 。在一个子带中只允许一个量化步长。C.熵编码 每个子带的小波分为矩形单元码块，它使用算术编码的方式（称为EBCOT）单独进行编码。这种截取降低了硬件和软件的储存空间的要求，并一定程度上提供了随机存取空间的位流。该块在所有子带中大小相同，使区块较低分辨率子带跨度较大的区域在原始图像。按一定的比率在每个子带中选择码块较大的矩形区，叫作辖区。三. JPEG2000功能 JPEG2000标准提供了大量的功能，其中包括感兴趣区域编码/解码，分辨率和质量的可伸缩性，和错误复原。A.感兴趣区域（ROI） 感兴趣区域的功能是允许编码或解码一副指定的图像中某区域图像的质量高于其余（背景）区域。这是由系数扩大影响感兴趣区域，使焦点与相关的感兴趣区域被放置在较高位的平面。在嵌入式编码过程中，这些位放置在非感兴趣区域图像前的位流。感兴趣区域编码方法中所定义的JPEG2000第1部分用处是通过规模变换完成，例如说，有些平面的感兴趣区域与其余的图像是完全分开的。这是以参考MAXSHIFT为形式，允许感兴趣区域编码的任意形状区域，而不需要形状的信息。B.可扩展性 新标准强调可扩展性的图像表达。部分压缩码流可提取，独立减压和恢复图像中所减少的分辨率，降低在任何给定的分辨率下的质量（SNR），或在减少空间区域，得到理想的分辨率和质量的区域。JPEG2000还支持完全无损压缩图像而不失去可扩展性。这意味着应用当中在任何给定的空间区域上都能够逐步提取更高质量的图像，最终实现该区域的无损化。这些功能在用户对压缩图像的应用上提供了新的互动模式。JPEG2000中的小波变换和位平面编码算法通过充分的比特流，使得图像具有分辨率和质量级的可扩展性。C.错误复原 许多应用中需要以容易出错的通信渠道提供图像数据。典型的无线通信渠道产生随机和突发位错误。互联网通信由于信道堵塞很容易造成数据包损失。为了提高这些容易出错的渠道压缩传递的性能，JPEG2000标准添加了错误复原分析和工具。错误复原工具包括用分割符号限制错误传播和定位，终止算术编码来检测错误和各种同步标记。四 JPEG2000的应用标准JPEG2000更好地服务于市场和各范围内应用，列举如下： 消费者应用，如多媒体设备（数码相机，个人数字助理， 3G手机，彩色传真机，打印机，扫描仪等）； 客户机/服务器之间的通讯（例如，互联网，图像数据库，视频流，视频服务器等）； 军事/监视（例如，高清卫星图像，运动检测，网络分发和储存等）； 医疗图像，DICOM标准规格的医疗数据交换； 遥感； 高品质的基于帧的视频录制，编辑和存储； 数字影院； JPEG2000设计了许多ICER图像压缩格式，是用来传送从火星探测器上传来的图片； 世界气象组织给JPEG2000添加新的GRIB2文件格式。此文件结构是专为全球气象数据分布而设计的。JPEG2000压缩在GRIB2的执行，使文件大小减少了高达80 。 五. 结论 JPEG2000是新一代静止图像压缩标准，它提供了一个新的框架和一个综合性工具箱，以便于更好地解决人们对数据压缩的需求。它提供了广泛的功能，如无损和有损编码，嵌入式损耗到无损编码，分辨率和质量的可扩展性，高质量，高压缩效率，错误恢复能力和感兴趣区域（ ROI ）的编码。结果表明， JPEG2000标准确实优于建立图像压缩标准。总体而言，JPEG2000的标准提供了丰富的功能，一个非常有效的方式和范围内统一的算法。由于技术的迅猛发展，它的价格有时带来不少压力，但这不应视为弊端。六 致谢这项工作受欧洲社会基金和毕达哥拉斯方案（合同 第89188 ）的资助。摘自：中国图形图像参考文献1 ISO/IEC JTC1/SC29/WG1 (ITU-SG8) N2058, JPEG2000 is now an International Standard, 9 March 2001.2 D. S. Taubman and M. W. Marcellin, JPEG2000: Image Compression Fundamentals, Standards, and Practice, Boston, MA: Kluwer, 2002.3 T. Acharya and P.-S. Tsai, JPEG2000 Standard for Image Compression, J. Wiley & Sons, New Jersey, 2005.4 P. Schelkens, A. Munteanu, A. Tzannes and C. Brislawn, JPEG2000 t 10 - Volumetric Data Encoding, Proc. 2006 IEEE Int Symposium on Circuits and Systems (ISCAS2006), Kos, Greece, May 2006.5 V. Spiliotopoulos, N. D. Zervas, C. E. Androulidakis, G. Anagnostopoulos and S. Theoharis, Quantizing the 9/7 Daubechies Filter Coefficients for 2D DWT VLSI Implementations, Proc. 2002 14th Int. Conf. On Digital Processing (DSP2002),vo1.I,pp.227-232,Santorini,Greece,July 2002.6 J.-S. Chiang, Y.-S.Lin and C.-Y. Hsieh, Efficient Pass-Parallel Architecture for EBCOT in JPEG2000, Proc. IEEE Int. Symposium on Circuits and Systems (ISCAS2002), vol. 1, pp. 773-776, Scottsdale, Arizona, May 2002.7 D.Taubman, “software Architectures for JPEG2000”,Proc.2002 14th Int. Conf. on Digital Signal Processing (DSP2002),vol. I, pp. 197-200,Santorini,Greece, July 2002.8 A.N. Skodras, C.A. Christopoulos and T. Ebrahimi, The JPEG2000 Still Image Compression Standard, IEEE Signal Proc. Magazine, vol. 18, no. 5, pp. 36.58, Sep. 2001.9 M. Rabbani and R. Joshi, An overview of the JPEG2000 Still Image Compression Standard, Signal Processing: Image Communication vol. 17, no. 1, pp 3-48, 2002.10 A. Bilgin and M. Marcellin, JPEG2000 for Digital Cinema, Proc. 2006 IEEE Int. Symposium on Circuits and Systems (ISCAS2006), Kos, Greece, May 2006. 11 A. Crespo and H. Garcia-Molina, Archival Storage for Digital Libraries, Proc. 3rd ACM Conf. on Digital Libraties, pp. 69-78, 1998.附：英文原文JPEG2000 Image Coding System Theory and ApplicationsAbstractJPEG2000, the new standard for still image coding, provides a new framework and an integrated toolbox to better address increasing needs for compression. It offers a wide range of functionalities such as lossless and lossy coding, embedded lossy to lossless coding, progression by resolution and quality, high compression efficiency, error resilience and region-of-interest (ROI) coding. Comparative results have shown that JPEG2000 is indeed superior to established image compression standards. Overall, the JPEG2000 standard offers the richest set of features in a very efficient way and within a unified algorithm. The price ofthis is its additional complexity, but this should not be perceived as a disadvantage, as the technology evolves rapidly.I. INTRODUCTION The JPEG2000 international standard represents advances in image compression technology where the image coding system is optimized not only for efficiency but also for scalability and interoperability in network and mobile environments. Digital imaging has become an integral part of the Internet and JPEG2000 is a powerful new tool that provides power capabilities for designers and users of networked imaging applications .With the progress of multimedia and Internet applications, the needs and requirements for image coding technologies grew and evolved. In March 1997, a call for contributions was launched for the development of a new standard for the compression of still images. This effort produced the JPEG2000 image coding International Standard in December 2000 (ISO 15444 / ITU-T Recommendation T.800) . Part 1 of the standard describes the core coding system, consisting of a limited number of possible coding algorithms, in order to provide maximum interchange. Part 2 consists of optional technologies not required for all implementations. Evidently, images encoded with Part 2 technology will not be able to be decoded with Part 1 decoders. As an example, Part 2 includes variable DC offset, multiple component transformations, Trellis Coded Quantization, user defined wavelets, arbitrary wavelet decompositions, general scaling-based ROI coding, and advanced error resilience schemes (Fig. 1).It is actually a toolbox with technologies useful for various specialized applications. Part 3 defines motion JPEG2000 (MJ2 or MJP2) and is based on Part 1 of JPEG2000. MJ2 will be used in many different areas, as for example in applications where it is desired to have a single codec for both still pictures and motion sequences (which is a common feature of digital still cameras), or in applications where very high quality motion pictures are required (e.g. medical imaging and motion picture production),or in video applications in error prone environments (e.g. wireless and the Internet).JPEG2000 compressed image sequences, synchronized audio and metadata can be stored in the MJ2 file format. Motion JPEG2000 is also targeting interoperability with the JPEG2000 file format (JP2) and the MPEG-4 file format (MP4). A new ad hoc group was created in Oct. 2002 (the Motion JPEG2000 for Medical Imaging) for the promotion of the JPEG2000 standards in the medical community. Part 4 of the standard defines the conformance testing. Part 5 defines the reference software (high quality free software). Two reference software implementations do exist, namely, the JPEG2000 software in Java and the JasPer software in C. The Kakadu software is also available .Part 6 defines the compound image file format. Part 7, originally intended to produce a technical report with guidelines of minimum support function of Part 1, has been withdrawn. Part 8 addresses security issues such as authentication, data integrity, protection of copyright and intellectual property, privacy, and conditional access. Part 9 defines interactivity tools, APIs and protocols for the transmission of JPEG2000 images over a client/server environment. Part 10 extents the JPEG2000 algorithm to 3D image compression and floating point data .Part 11 defines a file format for image compression and transmission in wireless environments such as mobile, wireless LAN and Radio. Part 12 defines the common file format for MPEG-4 and Motion JPEG2000. Finally, the purpose of Part 13 is to define a normative entry level JPEG 2000 encoder providing one or more optional complete encoding paths that use various features defined in the JPEG2000 standards.JPEG2000 is inherently more complex than its predecessor, JPEG. Its reliance on the discrete wavelet transform (DWT) and coding of wavelet coefficients in blocks, together imply a significantly higher cost in memory consumption than the baseline JPEG algorithm. The embedded entropy coding algorithm is far from trivial to implement and the code-stream parsing rules involve considerable conceptual complexity. The most demanding parts of the JPEG2000 engine are the DWT and the entropy coding.Numerous architectures, both hardware and software, have been proposed so far to reduce the computational and memory requirements. The demanding architecture of the JPE2000 is anticipated by its coding efficiency, which is much higher than that of the other coding approaches or standards, both for the lossless and lossy cases. II. JPEG2000 COMPRESSION ARCHITECTUREFigure 1 shows an illustration of how an image is transformed through the different stages of JPEG2000 encoding process. It is worth mentioning that, unlike many coding schemes, the JPEG2000 compression can be both lossy and lossless. JPEG2000 also allows for image tiling. The term tiling refers to the partition of the original (source) image into rectangular non-overlapping blocks (tiles), which are compressed independently, as though they were entirely distinct images. Arbitrary tile sizes are allowed, up to and including the entire image (i.e. no tiling). Tiling reduces memory requirements and constitutes one of the methods for the efficient extraction of a region of the image.A. Wavelet TransformTile components are decomposed into different decomposition levels using a separable wavelet transform. A decomposition level is related to the next decomposition level by spatial powers of two. Part 1 of the standard supports dyadic decomposition only, as this appears to yield the best compression performance for natural images. The DWT can be irreversible or reversible. The irreversible transform in Part 1 of JPEG2000 is implemented by means of a 9-tap/7-tap filter bank. The reversible transformation in Part 1 of JPEG2000 is implemented by means of a 5-tap/3- tap filter bank. B. QuantizationAt the encoder, the scalar quantization operation maps a given coefficient value to a quantizer index, which is then encoded as part of the compressed bit stream. At the decoder, the quantizer index is decoded and converted into the corresponding quantized value. This operation is lossy, unless the quantization step size is 1 and the coefficients are integers, as produced by the reversible integer 5/3 filters. Each of the transform coefficients of the subband b is bit uniformly quantized to the value . One quantization step size per subband is allowed. C. Entropy CodingEach subband of the wavelet decomposition is divided up into rectangular blocks, called code-blocks, which are coded independently using arithmetic coding by means of an approach called EBCOT. Such a partitioning reduces memory requirements in both hardware and software implementations and provides a certain degree of spatial random access to the bit-stream. The block size is identical for all subbands, so that the blocks in lower resolution subbands span a larger region in the original image. A neighborhood of spatially consistent code-blocks from each subband at a given resolution level forms larger rectangles, called precincts. III. JPEG2000 FUNCTIONALITIESThe JPEG2000 standard exhibits a large number of functionalities (features), among which one can mention regions of interest coding/decoding, resolution and quality scalability, and error resiliency.A. Region-of-Interest (ROI)ROI is a feature allowing for coding or decoding of certain regions in an image with better quality when compared to the rest (background) of the image. This is performed by scaling up coefficients influencing the ROI so that the bits associated with that ROI are placed in higher bit planes. During the embedded coding process, those bits are placed in the bit-stream before the non-ROI parts of the image. The ROI approach defined in the JPEG2000 Part 1 performs scale up shifts by values such that the bit planes for ROI are completely separate from those for the rest of the image. This is referred to as MAXSHIFT and allows ROI coding of arbitrary shaped regions without the need for shape information.B. ScalabilityThe new standard emphasizes scalable image representations. Portions of the compressed code-stream may be extracted and decompressed independently, to recover the image at a reduced resolution, at a reduced quality (SNR) within any given resolution, or within a reduced spatial region, at the desired resolution and quality. JPEG2000 also supports entirely lossless compression of images without sacrificing scalability. This means that an application is able to extract progressively higher quality representations of any given spatial region, leading eventually to a lossless representation of that region. These features provide applications and users with new paradigms for interacting with compressed imagery. The wavelet transform and bit plane coding in JPEG2000 algorithm allow for resolution and quality scalability types by an adequate construction of the bit-stream.C. Error Resilience Many applications require the delivery of image data over error prone communication channels. Typical wireless communication channels give rise to random and burst bit errors. Internet communications are prone to packet losses due to traffic congestion. To improve the performance of transmitting compressed images over these error prone channels, error resilient bitstream syntax and tools are included in the JPEG2000 standard. The error resilience tools include segmentation symbols to confine error propagation and localization, predictable termination in arithmetic coding to detect errors, and various synchronization markers.IV. Applications of JPEG 2000The markets and applications better served by this standard are listed below: Consumer applications such as multimedia devices (e.g., digital cameras, personal digital assistants, 3G mobile phones, color facsimile, printers, scanners, etc.) Client/server communication (e.g., the Internet, Image database, Video streaming, video server, etc.) Military/surveillance (e.g., HD satellite images, Motion detection, network distribution and storage, etc.) Medical imagery,the DICOM specifications for medical data interchange. Remote sensing High-quality frame-based video recording, editing and storage. Digital cinema JPEG 2000 has many design commonalities with the ICER image compression format that is used to send images back from the Mars rovers. World Meteorological Organization has built JPEG 2000 Compression into the new GRIB2 file format. The GRIB file structure is designed for global distribution of meteorological data. The implementation of JPEG 2000 compression in GRIB2 has reduced file sizes up to 80%.V. CONCLUSIONSJPEG2000, the new standard for still image coding, provides a new framework and an integrated toolbox to better address increasing needs for compression. It offers a wide range of functionalities such as lossless and lossy coding, embedded lossy to lossless coding, progression by resolution and quality, high compression efficiency, error resilience and region-of-interest coding. Comparative results have shown that JPEG2000 is indeed superior to established image compression standards. Overall, the JPEG2000 standard offers the richest set of features in a very efficient way and within a unified algorithm. The price of its additional