论文部分内容阅读
引言 本文介绍了两种实现二维付里叶变换的程序设计方法:第一种方法是直接实现二维变换,它适用于二维数组存放在内存贮器时的情况,算法完全可采用众所皆知的一维付里叶变换算法结构;第二种方法适用于二维数组存放在外存贮器时的情况,算法采用二次一维付里叶变换,其中一个方向很容易实现,另一个方向如果采用通常的一维变换程序,则二维数组需要转置,这样做比较费时。将点运算改为向量运算就可以不改变外存数据结构,同时实现这个方向上的一维变换,从而省去了转置过程。
Introduction This paper introduces two programming methods to achieve two-dimensional Fourier transform: The first method is the direct realization of two-dimensional transform, which is suitable for two-dimensional array stored in memory when the situation, the algorithm can be fully public The well-known one-dimensional Fourier transform algorithm structure; the second method is suitable for two-dimensional array stored in the external memory situation, the algorithm uses two-dimensional one-dimensional Fourier transform, one of the directions is easy to implement, and the other If we adopt the usual one-dimensional transform program in one direction, the two-dimensional array needs to be transposed, which is time consuming. Converting a point operation to a vector operation can not change the data structure of the external memory and at the same time realize a one-dimensional transformation in this direction, thereby eliminating the transpose process.