首页> 外文学位 >Digital image coding techniques: Codec design using vector quantization method.
【24h】

Digital image coding techniques: Codec design using vector quantization method.

机译:数字图像编码技术:使用矢量量化方法的编解码器设计。

获取原文
获取原文并翻译 | 示例

摘要

Recently there have been many efforts to improve coding systems for digital images and speech signals. Among the various techniques, a vector quantization (VQ) technique has proven that it can achieve a high compression ratio without sacrificing subjective and objective quality of an image.; During the past few years several design algorithms have been developed for a variety of vector quantizers and the performance of coders has been studied. The purpose of this research is to analyze some of these design techniques and facilitate their applications to digital image coding. The goal of this research project targeted and achieved a lower bit rate while maintaining image quality. Artificial neural networks such as Kohonen's self-organizing feature map (KSFM) and its variant have been applied and developed during this research. Several new encoding and decoding (codec) processes have been developed to maintain a high compression ratio (from 8 bits/pixel to 0.28 bits/pixel in test image Lena) without losing subjective and objective quality. I have obtained several experimental results which demonstrate the improvement of the new vector quantizers, making them comparable to industry standard and existing vector quantizers.; In recent years, the demand for digital video transmission and storage has increased dramatically in applications to medical images, teleconferencing (CCITT H.261), multi-media systems (MPEG-I, MPEG-II), and HDTV. In order to minimize the memory for storage and the bandwidth for transmission, digital video compression techniques with low implementation complexity have become crucial and mandatory. For digital video coding, a new vector quantization scheme (MCVQ) is proposed for interframe predictive coding of images to achieve a high compression ratio. Quantization method consists of a series of pipe-line processes. For frame sequences, motion compensation is used to reduce the variance of input vectors. Several codec simulations have been implemented and compared with industry standard such as MPEG and CCITT H.261. In this research a difference vector quantizer and subband coding (SBC) techniques (for the second layer of implementation) have been applied. The simulations show satisfactory results in implementation complexity and high compression ratio (from 60.7 Mbits/second to 1.2 Mbits/second) for video frame sequences.
机译:近来,已经进行了许多努力来改进用于数字图像和语音信号的编码系统。在各种技术中,矢量量化(VQ)技术已证明可以在不牺牲图像主观和客观质量的情况下实现高压缩比。在过去的几年中,已经为多种矢量量化器开发了几种设计算法,并且已经研究了编码器的性能。这项研究的目的是分析其中一些设计技术,并促进其在数字图像编码中的应用。该研究项目的目标是在保持图像质量的同时瞄准并实现了较低的比特率。在此研究期间,已应用和开发了诸如Kohonen的自组织特征图(KSFM)及其变体之类的人工神经网络。已经开发了几种新的编码和解码(codec)过程,以保持较高的压缩率(测试图像Lena中从8位/像素到0.28位/像素),而不会失去主观和客观质量。我已经获得了一些实验结果,这些结果证明了新矢量量化器的改进,使其可与行业标准和现有矢量量化器相媲美。近年来,在医疗图像,电话会议(CCITT H.261),多媒体系统(MPEG-I,MPEG-II)和HDTV的应用中,对数字视频传输和存储的需求急剧增加。为了最小化用于存储的存储器和用于传输的带宽,具有低实现复杂度的数字视频压缩技术已经变得至关重要和强制性。对于数字视频编码,提出了一种新的矢量量化方案(MCVQ),用于图像的帧间预测编码,以实现较高的压缩率。量化方法包括一系列管道过程。对于帧序列,使用运动补偿来减少输入向量的方差。已经实现了几种编解码器仿真,并将其与行业标准(例如MPEG和CCITT H.261)进行了比较。在这项研究中,已经应用了差异矢量量化器和子带编码(SBC)技术(用于第二层实现)。对于视频帧序列,仿真显示了在实现复杂性和高压缩率(从60.7 Mbits /秒到1.2 Mbits /秒)方面令人满意的结果。

著录项

  • 作者

    Shin, Yong Ho.;

  • 作者单位

    Kent State University.;

  • 授予单位 Kent State University.;
  • 学科 Computer Science.
  • 学位 Ph.D.
  • 年度 1994
  • 页码 122 p.
  • 总页数 122
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 自动化技术、计算机技术;
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号