Computational framework for family of order statistic filters for tensor valued data

Authors:
Bogusław Cyganek
Affiliations:
AGH – University of Science and Technology, Kraków, Poland
Venue:
ICIAR'06 Proceedings of the Third international conference on Image Analysis and Recognition - Volume Part I
Year:
2006

Citing 6
Cited 1

Multidimensional Orientation Estimation with Applications to Texture Analysis and Optical Flow

IEEE Transactions on Pattern Analysis and Machine Intelligence
Handbook of image processing operators

Handbook of image processing operators
Nonlinear image processing

Nonlinear image processing
Signal Processing for Computer Vision

Signal Processing for Computer Vision
Nonlinear Filters for Image Processing

Nonlinear Filters for Image Processing
Tensor processing for texture and colour segmentation

SCIA'05 Proceedings of the 14th Scandinavian conference on Image Analysis

Circular Road Signs Recognition with Affine Moment Invariants and the Probabilistic Neural Classifier

ICANNGA '07 Proceedings of the 8th international conference on Adaptive and Natural Computing Algorithms, Part II

Quantified Score

Hi-index	0.00

Visualization

Abstract

Nonlinear digital filters based on the order statistic belong to the very powerful methods of image restoration. The well known is the median filter operating on scalar valued images. However, the operation of the median filter can be extended into multi-valued pixels, such as colour images. It appears that we can go even further and define such filters for tensor valued data. Such a median filter for tensor valued images was originally proposed by Welk et.al. [10]. In this paper we present a different approach to this concept: We propose the family of nonlinear order statistic filters operating on tensor valued data and provide the computational framework for their non-numerical implementation.