Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A maximum entropy approach to unsupervised mixed-pixel decomposition.

Lidan Miao1, Hairong Qi, Harold Szu

  • 1Advanced Imaging and Collaborative Information Processing Group, Department of Electrical and Computer Engineering, University of Tennessee, Knoxville, TN 37996, USA. lmiao1@utk.edu

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|April 5, 2007
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Mathematical Framework for Quantifying Delocalization in MALDI-MSI via a Composite Scoring Approach.

ACS measurement science au·2026
Same author

Broiler Mobility Assessment via a Semi-Supervised Deep Learning Model and Neo-Deep Sort Algorithm.

Animals : an open access journal from MDPI·2023
Same author

Topological Structure and Semantic Information Transfer Network for Cross-Scene Hyperspectral Image Classification.

IEEE transactions on neural networks and learning systems·2021
Same author

Non-Local Representation Based Mutual Affine-Transfer Network for Photorealistic Stylization.

IEEE transactions on pattern analysis and machine intelligence·2021
Same author

Characterizing Sounds of Different Sources in a Commercial Broiler House.

Animals : an open access journal from MDPI·2021
Same author

Safety critical event prediction through unified analysis of driver and vehicle volatilities: Application of deep learning methods.

Accident; analysis and prevention·2021

This study introduces a new gradient descent maximum entropy (GDME) method for spectral unmixing. GDME offers robust and effective subpixel component estimation, outperforming least-squares methods in noisy conditions.

Area of Science:

  • Remote Sensing
  • Image Analysis
  • Signal Processing

Background:

  • Mixed pixels in remote sensing require subpixel scale analysis.
  • Spectral unmixing algorithms aim to identify constituent components (endmembers) and their abundances.
  • Existing least-squares methods are sensitive to noise and computationally expensive due to non-negativity constraints.

Purpose of the Study:

  • To propose an unsupervised spectral unmixing method addressing limitations of current approaches.
  • To develop a robust and effective algorithm for estimating endmembers and abundances.
  • To improve accuracy in noisy data or when endmember signatures are similar.

Main Methods:

  • Developed the gradient descent maximum entropy (GDME) method.
  • Utilized the maximum entropy principle for robust decomposition.

Related Experiment Videos

  • Applied geometric interpretations to justify the maximum entropy approach.

    • Main Results:

    • GDME provides more accurate estimates than least-squares methods under noise or similar endmember signatures.
    • The method demonstrates effectiveness on both simulated and real multispectral and hyperspectral data.
    • Experimental results validate the robustness and accuracy of the proposed GDME approach.

    Conclusions:

    • The gradient descent maximum entropy (GDME) method is a promising alternative for spectral unmixing.
    • GDME offers improved performance in challenging scenarios common in remote sensing.
    • The proposed method enhances the capability of analyzing subpixel composition in multispectral and hyperspectral imagery.