The Design of Multidimensional FIR Digital Filters by Phase Correction.
TEXAS UNIV AT AUSTIN ELECTRONICS RESEARCH CENTER
Pagination or Media Count:
This report presents a new technique for the design of multidimensional FIR digital filters. Given the desired amplitude response, the truncation error of a low order approximation to a high order filter is iteratively decreased by varying the phase response. Convergence of the iterative algorithm is discussed. In general, the phase responses of the filters resulting from the algorithm are nonlinear. Pictures suffer phase distortion when filters with nonlinear phase characteristics are used in filtering. A measure of phase distortion in pictures is introduced so that the effects of nonlinear phase in filters may be quantified. Examples illustrating the use of the design technique and the measure of phase distortion are included. The results of this report are formulated in two dimensions however they are valid for filters and signals of any dimension.