A nonlinear de-mixing method for hyperspectral images considering spectral variability

A hyperspectral image and variability technology, applied in the field of non-linear unmixing of hyperspectral images, which can solve the problems of not taking into account the spectral variability of nonlinear scenes and the deviation of unmixing results.

Active Publication Date: 2018-12-18
FUDAN UNIV
View PDF3 Cites 17 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the above-mentioned algorithms only focus on the nonlinear unmixing problem, and do not take into account the spectral variability existing in nonlinear scenes, which will still cause large deviations in the actual unmixing results

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A nonlinear de-mixing method for hyperspectral images considering spectral variability
  • A nonlinear de-mixing method for hyperspectral images considering spectral variability
  • A nonlinear de-mixing method for hyperspectral images considering spectral variability

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0115] The specific implementation manners of the present invention will be described below using simulated data and actual remote sensing image data as examples.

[0116] The nonlinear unmixing algorithm for hyperspectral remote sensing images that considers spectral variability is denoted by UNSUSC-SV.

[0117] 1. Simulation data experiment

[0118] In this section, the UNSUSC-SV algorithm is compared with the bi-objective non-negative matrix factorization algorithm Bi-objective NMF[8], and the nonlinear unmixing algorithm ASSKNMF[9] based on the abundance-constrained kernel non-negative matrix factorization, to compare the unmixing performance , and at the same time, the method proposed by the present invention without smoothing constraints on abundance and spectral variation coefficients is denoted as UNSU-SV to investigate the effect of adding smoothing constraints. Using the spectral angular distance SAD (Spectral Angle Distance) of the end member, the root mean square ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention belongs to the technical field of remote sensing image processing, in particular to a nonlinear de-mixing method for hyperspectral images considering spectral variability. At first, original data are mapped to a high-dimensional characteristic space by a kernel method, and the spectral variation coefficient is considered in the high-dimensional space to perform linear unmixing; at the same time, according to the spatial continuity of object distribution, local smoothing constraint is added to abundance and coefficient of variation, which makes them have spatial smoothness. This method can be used for unsupervised nonlinear spectral unmixing in the presence of spectral variability in Hapke and GBM nonlinear mixing models. The invention can overcome the problem of spectral variability existing in different nonlinear mixing scenes and improve the precision of spectral unmixing, and has important significance in practical application.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image processing, and in particular relates to a hyperspectral image nonlinear unmixing method. Background technique [0002] Remote sensing technology is a new comprehensive technology developed in the 1960s. It is closely related to science and technology such as space, electron optics, computer, and geography. It is one of the most powerful technical means for studying the earth's resources and environment. Hyperspectral remote sensing is a multi-dimensional information acquisition technology that combines imaging technology with spectral technology. Its images usually contain hundreds of bands and have high spectral resolution. It can obtain rich spectral information while obtaining the spatial distribution of ground objects. It is widely used in many fields such as military reconnaissance, environmental monitoring and geological exploration [1]. However, due to the low spatial resolut...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06T5/00G06T5/50
CPCG06T5/002G06T5/50G06T2207/10036
Inventor 智通祥王斌
Owner FUDAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap