Unlock instant, AI-driven research and patent intelligence for your innovation.

Blind end member extraction and spectrum unmixing method based on non-negative sparse auto-encoder

A self-encoder, non-negative sparse technology, applied in biological neural network models, character and pattern recognition, computer components, etc., can solve the problems of lack of blind endmember extraction methods and spectral unmixing methods, and achieve noise reduction. small effect

Pending Publication Date: 2022-04-12
BEIJING INST OF ENVIRONMENTAL FEATURES
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

There is currently a lack of extraction methods and spectral unmixing methods for blind endmembers

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
  • Blind end member extraction and spectrum unmixing method based on non-negative sparse auto-encoder
  • Blind end member extraction and spectrum unmixing method based on non-negative sparse auto-encoder
  • Blind end member extraction and spectrum unmixing method based on non-negative sparse auto-encoder

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

[0027] The basic idea of ​​the invention will first be explained below.

[0028] Studies have shown that a signal can be sparsely represented by a large number of other observed signals. This theory is also applicable to hyperspectral remote sensing images: each pixel can be linearly expressed by training samples, and these training samples f...

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 relates to a blind end member extraction and spectrum unmixing method based on a non-negative sparse auto-encoder, and one specific implementation mode of the method comprises the steps: carrying out the wave band selection of an original hyperspectral image, and inputting the spectrum channel data of any pixel in the hyperspectral image after the wave band selection into the non-negative sparse auto-encoder; the non-negative sparse auto-encoder comprises an input layer, a plurality of hidden layers and an output layer; under the condition that the output data of the output layer can reproduce the spectral channel data, obtaining an activation value of each neuron of the last hidden layer in the plurality of hidden layers and a weight vector of each neuron of the last hidden layer and the output layer; and determining the weight vector of each neuron and the output layer as the spectral data of one blind end member, and determining the activation value of the neuron as the abundance corresponding to the blind end member, thereby realizing the spectral unmixing of any pixel. According to the embodiment, high-precision blind end member extraction and spectrum unmixing can be realized through the non-negative sparse auto-encoder.

Description

technical field [0001] The invention relates to the technical field of hyperspectral data processing, in particular to a method for extracting blind end elements and spectral unmixing based on a non-negative sparse autoencoder. Background technique [0002] Hyperspectral imagery refers to spectral resolution at 10 -2 Spectral images within the magnitude range of λ (λ is the wavelength), in hyperspectral images, general pixels (i.e. pixels) are mixed pixels, which include spectral data of various ground objects, so it can be judged that in the mixed pixel Contains several kinds of endmembers (one endmember corresponds to one kind of feature), and quantitatively describes these kinds of endmembers, such as calculating the area percentage of these kinds of endmembers in this pixel, that is, the abundance of endmembers, The above process is endmember extraction and spectral unmixing of mixed pixels. [0003] In practical applications, the spectral data of the above endmembers ...

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): G06V10/40G06N3/04G06N3/06
Inventor 刘畅王广平
Owner BEIJING INST OF ENVIRONMENTAL FEATURES