【摘要】 二维离散小波变换是小波处理图像的关键算法。随着人们对多媒体信息需求的日益增长,处理速度正变得越来越重要,传统的先行后列式的二维离散小波变换计算方法具有变换速度慢,不能充分利用硬件设备的缺点。讨论了一种基于小波提升算法的、行列并行的二维离散小波变换改进算法,该算法具有运算速度快、占用内存少的特点,特别适用于嵌入式系统的需求。同时,基于TMS320C6000 DSP平台讨论了它的软件实现方法,经比较可比传统计算方法提高效率20%左右。更多还原

【Abstract】 Two-dimension discrete wavelet transform is the key algorithm for wavelet image processing.Due to increasing demand on multimedia information,processing speed becomes much more critical.Traditional way of calculation is slow and causes inefficient utilization of hardware.Dicusses an optimized two-dimension discrete wavelet transform based on lifting scheme,which transforms column data and row data simultaneously.It is much faster and less memory demanding,which benefits realvant embedded systems.Besides,software implementation on TMS320C6000 DSP is given and some 20% efficiency is achieved compared with traditional method.更多还原