Abstract
The authors describe a dedicated processor that can convolve a circular FIR filter of diameter 16 pixels to an image of 512 pixels multiplied by 512 pixels within 0. 5 s. Hexagonal sampling and a hexagonal sectioning method are used to realize the FIR filter, and discrete cyclic convolution, is performed using the Fermat number transform. The restoration of a blurred image is presented to show the effectiveness of the processor.
Original language | English |
---|---|
Pages (from-to) | 565-568 |
Number of pages | 4 |
Journal | ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings |
Publication status | Published - Dec 1 1986 |
ASJC Scopus subject areas
- Software
- Signal Processing
- Electrical and Electronic Engineering