Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Monte carlo characteristics dimension reduction method for small-sample hyperspectral image

A hyperspectral image and feature dimensionality reduction technology, applied in the fields of instruments, character and pattern recognition, computer parts, etc. The effect of increasing reliability, increasing complexity, increasing convenience

Inactive Publication Date: 2012-09-12
HARBIN ENG UNIV
View PDF1 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

These feature selection algorithms select some feature bands that contribute greatly to the classification, reducing redundant features in high-dimensional data, but they cannot give the optimal number of feature selections. great difficulty and inconvenience

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
  • Monte carlo characteristics dimension reduction method for small-sample hyperspectral image
  • Monte carlo characteristics dimension reduction method for small-sample hyperspectral image
  • Monte carlo characteristics dimension reduction method for small-sample hyperspectral image

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] Concrete realization steps of the present invention are:

[0041] 1. Read in hyperspectral image sample data x i ∈ R N , where n is the number of sample pixels, each pixel has N band features, samples are divided into M categories, x i is a sample in the sample data set X, R N Represents the N-dimensional feature space, and the sample data set X is set to be divided into M subsets according to the category, Z={z 1 ,z 2 ,…,z M}, the number of samples in each category is {β i},i=1,2,...,M, where z i is the whole cluster of the i-th category in the sample X, and the sample label is the y i ∈{1,2,...,M}.

[0042] 2. Calculate the intra-class compactness and inter-class separation coefficients of the hyperspectral image sample data under the current sample label, where the intra-class compactness coefficient of the kth feature band of the sample is

[0043] Compactness Coefficient ( k ) = ...

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 aims at providing a monte carlo characteristics dimension reduction method for a small-sample hyperspectral image. The method comprises the following specific steps of 1, selecting characteristic parameters of monte carlo characteristic dimension reduction algorithm; 2, generating a random number of the monte carlo characteristic dimension reduction algorithm; and 3, stastically estimating monte carlo characteristic parameters. When the characteristic parameters are selected, the intra-class compactness and the inter-class separatability of the hyperspectral image are considered, and the further processing reliability of the data is improved; and at the same time, the method can self-adaptively give an optimum dimension reduction wave band quantity, the important characteristic wave band in an original hyperspectral image is selected for the post-processing of the image, and the convenience in processing the high-dimensional data can be improved.

Description

technical field [0001] The invention relates to a hyperspectral image classification method, in particular to a small-sample hyperspectral image Monte Carlo feature dimensionality reduction method. Background technique [0002] As a basic application of hyperspectral image processing, hyperspectral image classification has long been widely concerned by experts, scholars and engineers from various countries. With the continuous development of hyperspectral remote sensing technology, the spectral resolution of hyperspectral scanners continues to increase. Compared with multispectral remote sensing, hyperspectral images provide more spectral information on ground features, making it more accurate to distinguish ground features. . Its higher spectral resolution enhances the ability to distinguish the nuances of ground objects, but also brings the curse of dimensionality (Hughes phenomenon), which seriously affects the accuracy of hyperspectral image classification. The main re...

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
IPC IPC(8): G06K9/62
Inventor 赵春晖齐滨王立国王玉磊李晓慧
Owner HARBIN ENG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com