Fast 16-bit Fixed-Point 8x8 IDCT Approximations

Author

Liu, Lijie ; Tran, Trac D. ; Topiwala, Pankaj

Author_Institution

Dept. of Electr. & Comput. Eng., Johns Hopkins Univ., Baltimore, MD

fYear

2006

fDate

Oct. 29 2006-Nov. 1 2006

Firstpage

989

Lastpage

993

Abstract

In this paper, we propose two 16-bit fixed-point IDCT schemes, i.e., the rotation-based approximation and the lifting-based approximation. Both schemes can pass IEEE1180 tests and maintain satisfactory compatibility with the floating-point IDCT. The structures of these two schemes are derived from Loeffler´s plane rotation-based factorization of the IDCT matrix. Coefficients in the structures are approximated carefully by dyadic rationales, leading to the fast multiplierless implementation. The drifting performance in H.263+, MPEG-2 and MPEG-4 is also demonstrated.

Keywords

approximation theory; discrete cosine transforms; fixed point arithmetic; IDCT matrix; Loeffler plane rotation-based factorization; discrete cosine transform; dyadic rationales; fixed-point IDCT; floating-point IDCT; lifting-based approximation; rotation-based approximation; Approximation error; Compaction; Decoding; Discrete cosine transforms; Discrete transforms; Heart; MPEG 4 Standard; Sparse matrices; Standards development; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Signals, Systems and Computers, 2006. ACSSC '06. Fortieth Asilomar Conference on

Conference_Location

Pacific Grove, CA

ISSN

1058-6393

Print_ISBN

1-4244-0784-2

Electronic_ISBN

1058-6393

Type

conf

DOI

10.1109/ACSSC.2006.354900

Filename

4176710