Parallelogrammic neighborhood (PN) system for nonparametric MRF based texture synthesis

Authors:
Arnab Sinha;Ajay Kumar KKV;Sumana Gupta
Affiliations:
Department of Electrical Engineering, Indian Institute of Technology Kanpur, India;Department of Electrical Engineering, Indian Institute of Technology Kanpur, India;Department of Electrical Engineering, Indian Institute of Technology Kanpur, India
Venue:
ICIP'09 Proceedings of the 16th IEEE international conference on Image processing
Year:
2009

Citing 4
Cited 0

Fast texture synthesis using tree-structured vector quantization

Proceedings of the 27th annual conference on Computer graphics and interactive techniques
Discovering texture regularity as a higher-order correspondence problem

ECCV'06 Proceedings of the 9th European conference on Computer Vision - Volume Part II
Texture synthesis via a noncausal nonparametric multiscale Markov random field

IEEE Transactions on Image Processing
Texture synthesis: textons revisited

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we introduce a new neighborhood system based Nonparametric Markov Random Field (NMRF) model for texture synthesis. The objective of this work is to generate minimal neighborhood system, in MRF based texture synthesis algorithms. Compared to the earlier neighborhood system (circular), the proposed one is not isometric and has fewer number of pixels within the neighborhood.This reduction in the number of neighborhood pixels does not degrade the synthesis results compared to the existing neighborhood systems.Two popular texture synthesis algorithms (based on nonparametric MRF) have been tested with the new neighborhood system for two standard texture databases.The results establish the efficacy of the proposed neighborhood system in terms of computational gain, with no visible degradation compared to the earlier systems.