Optimal difference operator selection

Authors:
Peter Veelaert;Kristof Teelen
Affiliations:
University College Ghent, Engineering Sciences, Ghent University Association, Ghent, Belgium;University College Ghent, Engineering Sciences, Ghent University Association, Ghent, Belgium
Venue:
DGCI'08 Proceedings of the 14th IAPR international conference on Discrete geometry for computer imagery
Year:
2008

Citing 6
Cited 1

A Discrete Expression of Canny's Criteria for Step Edge Detector Performances Evaluation

IEEE Transactions on Pattern Analysis and Machine Intelligence
The constuctive solution of linear systems of partial difference and differential equations with constant coefficients

Multidimensional Systems and Signal Processing
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision
Feature controlled adaptive difference operators

Discrete Applied Mathematics
Improving difference operators by local feature detection

DGCI'06 Proceedings of the 13th international conference on Discrete Geometry for Computer Imagery
Analysis and comparative evaluation of discrete tangent estimators

DGCI'05 Proceedings of the 12th international conference on Discrete Geometry for Computer Imagery

Adaptive and optimal difference operators in image processing

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

Differential operators are essential in many image processing applications. Previous work has shown how to compute derivatives more accurately by examining the image locally, and by applying a difference operator which is optimal for each pixel neighborhood. The proposed technique avoids the explicit computation of fitting functions, and replaces the function fitting process by a function classification process. This paper introduces a new criterion to select the best function class and the best template size so that the optimal difference operator is applied to a given digitized function. An evaluation of the performance of the selection criterion for the computation of the Laplacian for digitized functions shows better results when compared to our previous method and the widely used Laplacian operator.