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

K-means Hyperspectral Image Band Clustering Method Based on Mutual Information

A technology of hyperspectral image and clustering method, which is applied in the dimensionality reduction of hyperspectral image data and the field of remote sensing image processing, which can solve the problems of not being high, the separability of band aggregation categories, etc., and achieve the effect of high image classification accuracy

Active Publication Date: 2021-10-15
ZHEJIANG UNIV OF TECH
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the shortcomings of the existing unsupervised band selection algorithm that the bands are easy to gather in a certain band interval and the class separability is not high, the present invention provides a K-means hyperspectral based on mutual information with high class separability Image Band Clustering Method

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
  • K-means Hyperspectral Image Band Clustering Method Based on Mutual Information
  • K-means Hyperspectral Image Band Clustering Method Based on Mutual Information
  • K-means Hyperspectral Image Band Clustering Method Based on Mutual Information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0041] The present invention will be further described below in conjunction with the accompanying drawings.

[0042] refer to Figure 1 ~ Figure 3 , a K-means hyperspectral image band clustering method based on mutual information, the cluster center is selected through the average mutual information between bands and added to the K-means clustering iteration;

[0043] Different training samples and different classifiers are used to classify the selected representative bands to verify the effectiveness and applicability of the present invention.

[0044] Such as figure 1 Shown is the flowchart of the invented K-means hyperspectral image band clustering method based on mutual information, including the following steps:

[0045] 1) Use the multibandread() function in MATLAB to read the denoised and quantized hyperspectral image{B 1 ,...,B l ,...B L}, to determine the number k of bands to be selected. refer to figure 2 , the input image of the present invention is an India...

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

A K-means hyperspectral image band clustering method based on mutual information. The cluster center is selected through the average mutual information between bands and added to the K-means clustering iteration, including the calculation of the average mutual information, the determination of the cluster center, and the representative Selection, classification and evaluation of sex band combinations. By calculating the average mutual information inside and outside the band interval, the band is directly selected as the clustering center and added to the clustering iteration, and the K-means clustering is used to complete the clustering process of the band, so that the clustering of each band is representative and the image information is relatively large. Around the abundant bands, each cluster center band is used as the optimal band combination. The invention obtains higher image classification accuracy while retaining the spectral information of the hyperspectral image.

Description

technical field [0001] The invention belongs to the field of remote sensing image processing, and mainly relates to a method for unsupervised band selection of hyperspectral remote sensing images, which can be applied to the fields of dimensionality reduction, classification and target recognition of hyperspectral image data. Background technique [0002] With the development of spectral imager, hyperspectral remote sensing technology has become one of the hotspots in the field of remote sensing. Hyperspectral image refers to the continuous spectral image of the target ground object obtained with high spectral resolution in the spectral range from visible light to infrared light. Because of its rich spectral information, hyperspectral remote sensing has been widely used in environmental monitoring, target recognition and object classification. However, the rich spectral information of hyperspectral images comes at the cost of higher data dimension and larger data volume, wh...

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): G06K9/62G06K9/00
CPCG06V20/13G06V20/194G06F18/23213G06F18/2431
Inventor 孔燕萍覃亚丽
Owner ZHEJIANG UNIV OF TECH
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