A pan-sharpening algorithm based on joint sparsity

Xiao Xiang Zhu, Sofya Spiridonova, Richard Bamler

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

12 Scopus citations

Abstract

Recently sparse signal representation of image patches was explored to solve the pan-sharpening problem. Although the proposed sparse reconstruction based methods lead to motivating results, yet none of them has considered the fact that the information contained in different multispectral channels may be mutually correlated. In this paper, we extend the Sparse Fusion of Images (SparseFI, pronounced "sparsify") algorithm, proposed by the authors before, to a Jointly Sparse Fusion of Images (J-SparseFI) algorithm by exploiting these possible signal structural correlations between different multispectral channels. This is done by making use of the distributed compressive sensing (DCS) theory that restricts the solution of an underdetermined system by considering an ensemble of signals being jointly sparse. The algorithm is validated with UltraCam data.

Original languageEnglish
Title of host publicationProceedings of the 2012 Tyrrhenian Workshop on Advances in Radar and Remote Sensing
Subtitle of host publicationFrom Earth Observation to Homeland Security, TyWRRS 2012
Pages177-184
Number of pages8
DOIs
StatePublished - 2012
Event2012 Tyrrhenian Workshop on Advances in Radar and Remote Sensing: From Earth Observation to Homeland Security, TyWRRS 2012 - Naples, Italy
Duration: 12 Sep 201214 Sep 2012

Publication series

NameProceedings of the 2012 Tyrrhenian Workshop on Advances in Radar and Remote Sensing: From Earth Observation to Homeland Security, TyWRRS 2012

Conference

Conference2012 Tyrrhenian Workshop on Advances in Radar and Remote Sensing: From Earth Observation to Homeland Security, TyWRRS 2012
Country/TerritoryItaly
CityNaples
Period12/09/1214/09/12

Fingerprint

Dive into the research topics of 'A pan-sharpening algorithm based on joint sparsity'. Together they form a unique fingerprint.

Cite this