A new kind of nonlinear filter, the piecewise-quadratic (PWQ) filter, is presented. This kind of filter is based on canonical PWQ functions, and enables filtering via domain partition and local processing which is particularly suitable for processing signals with various statistical properties such as images. The application of such a filter in restoring noisy images with Gaussian and uniform noise distributions is shown.
References
-
-
1)
-
W. Li ,
Ji-Nan Lin ,
R. Unbehauen
.
Canonical representation of piecewise-polynomial functions with nondegeneratelinear domain partition.
IEEE Trans. Circuits Syst.-I
-
2)
-
Ji-Nan Lin ,
R. Unbehauen
.
Adaptive nonlinear digital filter with canonical piecewise-linear structure.
IEEE Trans. Circuits Syst.
,
347 -
353
-
3)
-
V.J. Mathews
.
Adaptive polynomial filters.
IEEE SP Mag.
,
3 ,
10 -
26
-
4)
-
R.P. Lippmann
.
An introduction to computing with neural nets.
IEEE ASSP Mag.
,
4 -
22
-
5)
-
J.-N. Lin ,
R. Unbehauen
.
Canonical piecewise-linear approximations.
IEEE Trans. Circuits Syst. I
,
697 -
699
http://iet.metastore.ingenta.com/content/journals/10.1049/el_19981190
Related content
content/journals/10.1049/el_19981190
pub_keyword,iet_inspecKeyword,pub_concept
6
6