Approximate convolution using DCT coefficient multipliers

Author

Merhav, Neri ; Kresch, Renato

Author_Institution

Dept. of Electr. Eng., Technion-Israel Inst. of Technol., Haifa, Israel

Volume

8

Issue

4

fYear

1998

fDate

8/1/1998 12:00:00 AM

Firstpage

378

Lastpage

385

Abstract

We develop a method for designing discrete cosine transform (DCT) coefficient multipliers in order to approximate the operation of two-dimensional (2-D) convolution of an image with a given kernel. The method is easy to implement on compressed formats of DCT-based compression methods (JPEG, MPEG, H.261) by using decoding quantization tables that are different from the encoding quantization tables

Keywords

approximation theory; convolution; decoding; discrete cosine transforms; filtering theory; image coding; image enhancement; multiplying circuits; quantisation (signal); transform coding; two-dimensional digital filters; 2D convolution; 2D linear filtering; DCT coefficient multipliers; DCT-based compression methods; H.261; JPEG; MPEG; approximate convolution; compressed formats; decoding quantization tables; discrete cosine transform; encoding quantization tables; image coding; image enhancement; Convolution; Decoding; Design methodology; Discrete cosine transforms; Encoding; Image coding; Kernel; Quantization; Transform coding; Two dimensional displays;

fLanguage

English

Journal_Title

Circuits and Systems for Video Technology, IEEE Transactions on

Publisher

ieee

ISSN

1051-8215

Type

jour

DOI

10.1109/76.709404

Filename

709404