【作者】 安丹丹；

【导师】 王宝珠；

【作者基本信息】 河北工业大学 ， 通信与信息系统， 2007， 硕士

【摘要】 随着Internet网络、无线通信技术以及交互式多媒体技术的迅速发展,视频压缩技术成为当今视频通信的研究重点。在最近的二十多年中,小波理论得到了迅速的发展,成功地在许多领域得到应用。在图像压缩领域,基于小波变换的JPEG2000已经成为当今主流的图像压缩标准。在视频压缩领域,基于小波的压缩算法不断出现,并且表现出其优异的性能和良好的发展前景。本文以此为背景,提出基于小波变换的视频压缩算法研究,首先对小波变换理论进行了较为深入的阐述,包括多分辨率分析理论、图像小波分解过程的推导等,对图像分解中小波基的选择问题进行了讨论,从而确定了课题的实施方案,即在本文设计的视频压缩算法中采用双正交小波基D9/7。其次,研究了基于小波变换的静止图像压缩算法EZW、SPIHT和一种改进的EZW算法,这些算法是视频压缩编码算法中的关键部分。在软件平台上进行了算法验证、分析和对比,结果显示,应用改进的EZW算法对图象进行压缩,重构图像的PSNR值较高。论文第三部分研究了基于小波变换的视频压缩算法和多分辨率运动补偿算法,对运动补偿算法进行了适当改进。最后,设计了一个基于小波变换、小波域运动补偿和改进的EZW算法的视频压缩方案,进行了软件设计并对其运行结果进行了数据分析。经与H.263视频压缩标准对比,说明本文提出的算法在中高码率下性能优异,在低码率下也达到了H.263标准。更多还原

【Abstract】 With the rapid development of Internet network, wireless communication, and interactive multimedia, video compression is becoming an important issue in the research area of video communication today. In the last twenty more years, wavelet theory is developed rapidly , and the research areas of wavelet theory are also expanded thoroughly. Wavelet theory has been applied to many fields successfully. In the field of still-image compression, the JPEG2000, which based on wavelet transform has become the most popular standard of image compression. In the video compression field, new compression algorithm based on wavelet appears and exhibit excellent performance and good future.Under this background, this paper raises the research on video compression algorithm using wavelet transform.Firstly, wavelet theory is dedply elaborated,including MRME , derivation of image deposition using wavelet and the problem of wavelet bases selection ,all of this determines the program of the paper.Secondly,studies the methods of image compresstion EZW,SPITH and an improved EZW,and displays the experiment result.The result shows that image compression using improved EZW,the reconsruction imgae gets higher PSNR.Then reaserch the video compression algorithm and MRME ,improve the lgorithm.Finally,this paper designs wavelet video compression systerm and displays the results.The results shows that the algorithm is merit under high rate,and reach the H.263 under low rate.更多还原