基于小波变换的人脸图像压缩-模式识别与智能系统专业论文.docxVIP

摘要由于互联网络，无线网络等传输媒介的发展以及人们的交流信息要求对多媒 摘要 由于互联网络，无线网络等传输媒介的发展以及人们的交流信息要求对多媒 体应用的要求的日益增多，并且随着数学工具以及信号处理技术的不断发展，图 像压缩技术在应用上正在朝着低存储空间，高压缩率，编解码快速，能够进行多 分辨率分析，能够进行渐进传输以及在特定应用下的高度适应性。作为人类视觉 的重心之一，尤其是在数字图片广泛普及之后，人脸图像的出现是很频繁的，所 以在可预计的将来，在人脸库，可视电话，门禁系统等需要低码率以及甚低码率 情况下有广泛的应用前景。 本文对人类视觉系统进行了仔细的探讨，分析其发展过程和趋势，并找出其 对于人脸图像压缩的指导意义，并且在详细探讨了图像压缩的评价标准之后提出 了一种基于主观评价的图像评价标准。 本文详细的分析了图像压缩算法的发展并且仔细地比较了他们的特点并选 择一种比较适宜本文目的的算法进行改进。在此基础上，在根据强边，纹理和平 滑区域的主观视觉分类下结合人脸固有的先验知识对人脸图像进行处理。 本文提出了SPIHT算法下的剪枝补枝的思想，并且按照其在人类视觉系统的 指导意义下赋予一定的权值，分配到不同的位平面，并按照优先级进行编码和解 码进行渐进输出。 关键词： 人类视觉系统SPIHT算法人脸图像压缩人脸特征 ABSTRACTWith ABSTRACT With the development of internet，wireless net，multimedia technique， mathematical tools，and signal processing technology,image compression technology in the application are moving towards the quality of low storage space，hil曲 compression rate，encoding fast，conducting multi·resolution analysis，being capable of progressive transmission，and in particular having highly adaptive application．As the focus of the human visual system，especially in the widely popular digital photographs，the appearance of the facial images is very frequent．Therefore，in the foreseeable future，in the face database，video phone，access control system，and other aspect that acquiring low bit rate transmission there broad application prospects． In this paper,human visual system has been carefully discussed and its development is carefully analyzed．Then new judging method about image quality is presented based on subjective judgment． A detailed analysis and comparison of their features of the image compression algorithm is presented to improve the algorithm according to the appropriate aims． EspeciaUy about the course of its development and trends，and its significance is identified．According to the strong edge，texture and smooth region of visual classification and under the facial prior knowledge the inherent right facial image is processed． This paper presents pruning and filling the branches algorithm under the SPIHT in acco