Hyperspectral Image Compressive Sensing Method Based on Reweighted Laplacian Sparse Prior

A hyperspectral image and sparse prior technology, which is applied in the field of hyperspectral image compression sensing based on reweighted Laplacian sparse prior, can solve the problem of low reconstruction accuracy

Active Publication Date: 2017-11-24
NORTHWESTERN POLYTECHNICAL UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the deficiency of low reconstruction accuracy of existing hyperspectral image compression sensing methods, the present invention provides a hyperspectral image compression sensing method based on reweighted Laplacian sparse prior

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
  • Hyperspectral Image Compressive Sensing Method Based on Reweighted Laplacian Sparse Prior
  • Hyperspectral Image Compressive Sensing Method Based on Reweighted Laplacian Sparse Prior
  • Hyperspectral Image Compressive Sensing Method Based on Reweighted Laplacian Sparse Prior

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0060] The specific steps of the hyperspectral image compressed sensing method based on the heavy-weighted Laplacian sparse prior of the present invention are as follows:

[0061] Will contain n b Bands, each band contains n p Each band of the hyperspectral image of pixels is stretched into a row vector, and all row vectors form a two-dimensional matrix. Among them, each column of X represents the spectrum corresponding to each pixel, and this direction is the spectral dimension; each row of X corresponds to all the pixel values ​​of a band, and this direction is the spatial dimension. The present invention mainly includes the following four steps, which are specifically as follows:

[0062] 1. Obtain compressed data.

[0063] Gaussian random observation matrix with column normalization Sampling the spectral dimension of hyperspectral image X to obtain compressed data m b Indicates the length of the band after compression, ρ=m b / n b Is the sampling rate.

[0064] G=AX+N (1) where,...

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 discloses a hyperspectral image compression sensing method based on re-weighted Laplace sparse prior, which is used to solve the technical problem of low reconstruction accuracy of the existing hyperspectral image compression sensing method. The technical solution is to randomly collect a small number of linear observations of each pixel spectrum as compressed data, establish a compressed sensing model based on reweighted Laplace sparse prior and a sparse regularized regression model, and solve the established model. Since a small number of linear observations are randomly collected as compressed data, resource consumption during image acquisition is reduced. The re-weighted Laplacian sparse prior accurately describes the strong sparsity in hyperspectral images, overcomes the non-uniform constraints of traditional Laplacian sparse priors on non-zero elements, and improves the reconstruction accuracy of hyperspectral images. After testing, when the sampling rate is 0.15 and there is strong noise with a signal-to-noise ratio of 10db in the compressed data, the peak signal-to-noise ratio of the present invention is improved by more than 4db compared with the method of the background technology.

Description

Technical field [0001] The present invention relates to a hyperspectral image compressed sensing method, in particular to a hyperspectral image compressed sensing method based on a heavy-weighted Laplacian sparse prior. Background technique [0002] The spectral information of hyperspectral images is helpful for the detection, positioning, classification and identification of remote sensing features. However, the huge data volume of hyperspectral images puts forward strict requirements on the soft and hard resources in image collection, transmission and processing, which restricts hyperspectral images. Image application. Therefore, the hyperspectral image compression algorithm is one of the research hotspots in the hyperspectral field. At present, a large number of common image compression methods have been successfully applied to hyperspectral images. However, this type of method can only compress the acquired images, and cannot reduce the huge resource demand in the imaging p...

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 Patents(China)
IPC IPC(8): H03M7/30
Inventor 魏巍张艳宁张磊严杭琦
Owner NORTHWESTERN POLYTECHNICAL 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