Image Processing and the Arithmetic Fourier Trans-form

A new Fourier technique, the Arithmetic Fourier Transform (AFT) was recently developed for signal processing. This approach is based on the number-theoretic method of Mobius inversion. The AFT needs only additions except for a small amount of multiplications by prescribed scale factors. This new algorithm is also well suited to parallel processing. And there is no accumulation of rounding errors in the AFT algorithm. In this paper, the AFT is used to compute the discrete cosine transform and is also extended to 2-D cases for image processing. A 2-D Mobius inversion formula is proved. It is then applied to the computation of Fourier coefficients of a periodic 2-D function. It is shown that the output of an array of delay-line (or transversal) filters is the Mobius transform of the input harmonic terms. The 2-D Fourier coefficients can therefore be obtained through Mobius inversion of the output the filter array.

[1]  D W. Tufts,et al.  Arithmetic Fourier Transform And Adaptive Delta Modulation: A Symbiosis For High Speed Computation , 1988, Photonics West - Lasers and Applications in Science and Engineering.

[2]  William K. Pratt,et al.  Scene Adaptive Coder , 1984, IEEE Trans. Commun..

[3]  Anil K. Jain,et al.  A Sinusoidal Family of Unitary Transforms , 1979, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Robert M. Harlick.  A Storage Efficient Way to Implement the Discrete Cosine Transform , 1976, IEEE Trans. Computers.

[5]  Wen-Hsiung Chen,et al.  A Fast Computational Algorithm for the Discrete Cosine Transform , 1977, IEEE Trans. Commun..

[6]  J. Pearl,et al.  Comparison of the cosine and Fourier transforms of Markov-1 signals , 1976 .

[7]  G. Sadasiv.  The arithmetic Fourier transform , 1988, IEEE ASSP Magazine.

[8]  C. Smith,et al.  Adaptive Coding of Monochrome and Color Images , 1977, IEEE Trans. Commun..