A unifying view for stack filter design based on graph search methods

Authors:
Nina S. T. Hirata;Junior Barrera
Affiliations:
Computer Science Department of Institute of Mathematics and Statistics, University of São Paulo, Rua do Matão, 1010 05508-090 São Paulo, SP, Brazil;Computer Science Department of Institute of Mathematics and Statistics, University of São Paulo, Rua do Matão, 1010 05508-090 São Paulo, SP, Brazil
Venue:
Pattern Recognition
Year:
2005

Citing 12
Cited 0

Minimal representations for translation-invariant set mappings by mathematical morphology

SIAM Journal on Applied Mathematics
Decomposition of mappings between complete lattices by mathematical morphology, part I.: General lattices

Signal Processing
Efficient derivation of the optimal mean-square binary morphological filter from the conditional expectation via a switching algorithm for discrete power-set lattice

Circuits, Systems, and Signal Processing
Automatic programming of morphological machines by PAC learning

Fundamenta Informaticae - Special issue on mathematical morphology
The weighted median filter

Communications of the ACM
Fast algorithms for training stack filters

IEEE Transactions on Signal Processing
Weighted order statistics filters-their classification, someproperties, and conversion algorithm

IEEE Transactions on Signal Processing
Finding of optimal stack filter by graphic searching methods

IEEE Transactions on Signal Processing
Design of optimal stack filters under the MAE criterion

IEEE Transactions on Signal Processing
Stack filter design: a structural approach

IEEE Transactions on Signal Processing
A training framework for stack and Boolean filtering-fast optimal design procedures and robustness case study

IEEE Transactions on Image Processing
A fast algorithm for designing stack filters

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.01

Visualization

Abstract

Stack filters are operators that commute with the thresholding operation, i.e., thresholding a signal, applying the binary filter on each thresholded binary signals, and then summing up (stacking) the results yields the same result as applying the multi-level (gray-scale) filter on the original signal. Several approaches for designing optimal stack filters from training data, where optimality is characterized in terms of costs based on input-output joint observations, have been proposed. This work considers stack filter design from training data under a general statistical framework developed in the context of morphological image operator design. This framework (1) provides a common point of view for distinct design approaches, being useful for comparative analysis or for emphasizing differences, (2) clearly answers the issue of why binary signals from different threshold levels, although following distinct distributions, can be pooled together in the cost estimation process, and (3) helps to show that several stack filter design approaches based on lattice diagrams search methods share a common underlying formulation.