The discrimination capability (DC) is an important quality parameter in pat
tern recognition. If the input scene suffers some degradation, this paramet
er decreases. A source of degradation could be the imaging system, in parti
cular the obtained image could be out of focus. In this paper we analyze si
ngle and multichannel methods to obtain a DC invariant to the defocus. They
are based on the Laplacian pyramidal decomposition of the filter. In the c
ase of multichannel procedures we analyze several nonlinear methods to comb
ine the information given by each channel. (C) 2000 Published by Elsevier S
cience B.V.