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

Hyperspectral remote sensing image classification method and system based on multi-saliency feature fusion

A hyperspectral remote sensing and feature fusion technology, which is applied in the field of hyperspectral remote sensing image classification methods and systems, can solve the problems of long training time and difficulty in parameter adjustment, so as to make up for the difficulty of model parameter adjustment, increase the leading role of classification, and improve the utilization of features rate effect

Active Publication Date: 2021-06-18
SHANDONG LANGCHAO YUNTOU INFORMATION TECH CO LTD
View PDF8 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problems of difficult parameter adjustment and long training time in the current popular deep neural network-based hyperspectral remote sensing image classification method, the present invention provides a hyperspectral remote sensing image classification method and system based on multi-significant feature fusion to improve Feature Utilization During Classification

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 remote sensing image classification method and system based on multi-saliency feature fusion
  • Hyperspectral remote sensing image classification method and system based on multi-saliency feature fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0053] combined with figure 1 , this embodiment proposes a hyperspectral remote sensing image classification method based on the fusion of multiple salient features, and its implementation includes:

[0054] Step S1, using principal component analysis (PCA) to reduce the dimensionality of the original hyperspectral remote sensing image.

[0055] Step S2, for the hyperspectral remote sensing image after dimensionality reduction, first use the extended morphology method EMP to obtain multiple morphological feature maps, and then use the local binary model LBP and gradient histogram HOG to extract the texture features and gradients of the morphological feature maps feature, and then use the saliency BMS based on Boolean mapping to obtain the saliency feature Ⅰ based on the texture feature and the saliency feature Ⅱ based on the gradient feature. The saliency feature Ⅰ and saliency feature Ⅱ are fused, and the corresponding pixel values ​​are bit by bit Take the average value to ...

Embodiment 2

[0073] This embodiment proposes a hyperspectral remote sensing image classification system based on the fusion of multiple salient features, which includes a dimensionality reduction module 1, an EMP module 2, an LBP module 3, a HOG module 4, a BMS module 5, a fusion module 6, and a fusion module 2.7. Fusion module 3.8. Random forest module 9. combined with figure 2 , in order to better describe the data transfer process between each module, attached figure 2 Two LBP modules 3 and two HOG modules 4 are drawn in the figure. In fact, there is only one LBP module 3 and one HOG module 4 in this system. Attached figure 2 The solid arrows and hollow arrows represent the different processing processes of the original hyperspectral remote sensing image, and the double hollow arrows represent the fusion feature C finally output by the fusion module 38.

[0074] Dimensionality reduction module 1 uses principal component analysis (PCA) to reduce the dimensionality of the original hy...

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 remote sensing image classification method and system based on multi-saliency feature fusion, and relates to the technical field of image classification, and the technical scheme is that the method comprises the steps of: carrying out the dimension reduction of an original hyperspectral remote sensing image; obtaining a morphological feature map of the dimension-reduced image by using an extended morphological method EMP; for the dimension-reduced image and the morphological feature map, extracting texture features and gradient features by using a local binary pattern (LBP) and a histogram of gradient (HOG), and obtaining texture and gradient-based saliency features by using a saliency BMS based on Boolean mapping; and carrying out operations of fusing the texture and gradient-based saliency features and then integrally fusing all the saliency features, and carrying out pixel point classification by using a random forest classification algorithm according to a final fusion result. According to the method, the feature utilization rate in the classification process can be improved, and the problems of difficulty in model parameter adjustment, easiness in overfitting, high training cost and the like in the current hyperspectral remote sensing image classification method based on the deep neural network are solved.

Description

technical field [0001] The invention relates to the technical field of image classification, in particular to a hyperspectral remote sensing image classification method and system based on fusion of multiple salient features. Background technique [0002] Hyperspectral remote sensing technology can simultaneously use dozens or even hundreds of extremely narrow spectral bands to collect data, capture more data and provide more valuable information. Hyperspectral remote sensing images collected by hyperspectral remote sensing technology are usually referred to as hyperspectral remote sensing images for short. The quality of remote sensing images is getting higher and higher. The utilization, processing and classification of hyperspectral remote sensing images are the key steps to realize the implementation of this industry and its value. [0003] Feature extraction is a key step in hyperspectral remote sensing image classification. If the original hyperspectral data is direct...

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): G06K9/62G06K9/46G06K9/00
CPCG06V20/13G06V10/50G06F18/213G06F18/24323G06F18/253Y02A40/10
Inventor 张睿智孙思清翟盛龙赵志航王东伟
Owner SHANDONG LANGCHAO YUNTOU INFORMATION TECH CO LTD
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