In this paper, the regularized constrained total least-squares technique an
d its applications to signal restoration are presented. An error analysis i
s also performed to ascertain its statistical characteristics. Results for
both simulated data at different noise levels and measured data with unknow
n noise characteristics are presented. A comparison with other restoration
techniques is also made to demonstrate the applicability of this technique.
(C) 1998 Elsevier Science B.V. All rights reserved.