【作者】 何耀华；

【导师】 赵亦工；

【作者基本信息】 西安电子科技大学 ， 模式识别与智能系统， 2002， 硕士

【摘要】 本文主要研究高保真静止图象压缩算法，包括嵌入式零树编码算法和JPEG2000。其中对嵌入式零树编码算法的研究有其现实意义和理论价值，SPIHT和LZC是较有影响的经典算法，但由于对存储空间的需求较大，在硬件实现上有一定的难度。本文在SPIHT的基础上提出了一种适于硬件实现的无列表零树编码算法，该算法结构简单，存储空间开销小，而且重建图象质量仅略逊于SPIHT。本文还对LZC进行了改进，使之通过使用移位和加操作省去两个状态表，节约了硬件资源。同时，本文还对静止图象压缩标准JPEG2000的核心算法进行了C语言实现，包括EBCOT算法、小波提升算法、MQ算术编码器等。更多还原

【Abstract】 This paper deals with the algorithms of still image compression with high fidelity, including embedded zerotree coding and JPEG2000. The study on the algorithm of embedded zerotree coding has applicable and theoretical value. SPIHT and LZC are two classical algorithms of it, which is difficult to hardware implementation for its large space of storage. Based on SPIHT, no lists zerotree coding algorithm (NLS) has been put forward to be applicable for hardware implementation. Simple in structure, small in storage space, the NLS algorithm is very close to SPIHT in the quality of reconstructing image. Some improvements have maded to save hardware resource by using shifting and adding operations which are instead of two static maps. The core algorithm of still image compression standard JPEG2000 are simulated with Watcom C, including the algorithm of EBCOT, the lifting wavelet transform and MQ arithmetic coder.更多还原