A Convex Model for Matrix Factorization and Dimensionality Reduction on Physical Space and Its Application to Blind Hyperspectral Unmixing

Abstract

A collaborative convex framework for factoring a data matrix X into a non-negative product AS, with a sparse coefficient matrix S, is introduced. We restrict the columns of the dictionary matrix A to coincide with certain columns of X, thereby guaranteeing a physically meaningful dictionary and dimensionality reduction. As an example, we show applications of the proposed framework on hyperspectral endmember and abundances identification.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 01, 2010
Accession Number
ADA540658

Entities

People

  • Ernie Esser
  • Guillermo Sapiro
  • Jack Xin
  • Michael Moeller
  • Stanley Osher

Organizations

  • University of Minnesota

Tags

Communities of Interest

  • Sensors

DTIC Thesaurus Topics

  • Algorithms
  • California
  • Computational Biology
  • Construction
  • Detection
  • Detectors
  • Dictionaries
  • Dimensionality Reduction
  • Engineering
  • Hyperspectral Imagery
  • Infrared Spectra
  • Materials
  • Mathematics
  • Sensor Networks
  • Tree Canopy
  • Universities
  • Urban Areas

Readers

  • Educational Psychology
  • Graph Algorithms and Convex Optimization.
  • Image Processing and Computer Vision.

Technology Areas

  • Space