Research Note
A Parallel Algorithm for 4×4 DCT

https://doi.org/10.1006/jpdc.1999.1532Get rights and content

Abstract

By developing a generalized 1D approach and parallel computing algorithm, this paper presents a parallel algorithm design and hardware implementation for the computation of 4×4 DCT. This algorithm sorts all the 2D input pixel data into four groups. Each group is then forwarded to a 1D DCT arithmetic unit to complete the computation. After a few simple additions which are designed to follow the output of 1D DCTs, the computation of 2D DCT is implemented in parallel. Therefore, the efficiency of the algorithm is entirely dependent on the 1D DCT algorithm adopted, and all the existing fast algorithms for 1D DCT can be directly applied to further optimise the algorithm design. The development can also be further extended to compute general 2D DCT by a recursive procedure where the 4×4 DCT algorithm is used as the basic core.

References (11)

  • K.R. Rao et al.

    Discrete Cosine Transform Algorithms, Advantages, and Applications

    (1990)
  • B.G. Lee

    A new algorithm to compute the discrete cosine transform

    IEEE Trans. Acoust. Speech, Signal Process.

    (1984)
  • J. Makhoul

    A fast cosine transform in one and two dimensions

    IEEE Trans. Acoust. Speech, Signal Process.

    (1980)
  • F. Ephraim

    Fast algorithms for the discrete cosine transform

    IEEE Trans. Signal Process.

    (1992)
  • H. J. Lee, J. B. Lee, H. Byun, Comparisons of fast 2D DCT algorithms for parallel programmable digital signal...
There are more references available in the full text version of this article.

Cited by (0)

f1

E-mail: [email protected]

View full text
Copyright © 1999 Academic Press. All rights reserved.