一维小波变换及其逆变换结构

小波变换是信号处理、图像压缩和模式识别等诸多领域中一个非常有效的数学分析工具．然而，实时小波变换计算量大，需要专用硬件来实现．在专用小波变换芯片设计中，小波变换及其逆变换的结构设计已经成为关键技术．本文提出用Systolic阵列结构来完成小波变换及其逆变换，利用缓冲器来解决小波逆变换重建延迟问题．仿真结果表明本文所提结构有效，含乘法器、寄存器数目较少．

A One-dimensional Forward and Inverse Wavelet Transform Architecture

The wavelet transform (WT) is a very effective mathematical tools in many fields such as signal processing, image coding and pattern recognition, but it needs a special-purpose hardware to implement its huge computation. In the design of the special-purpose WT chip,the key technique is to design the configurations of WT and the inverse wavelet transform (IWT). This paper proposed an Systolic array configuration, which can perform the WT and IWT by using buffer to resolve the reconstruction delay. The simulation results show that the Systolic array is feasible and has good properties such as minimum number of registers and multipliers.