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

A Semantic Annotation Method for Hyperspectral Remote Sensing Images

A technology of hyperspectral remote sensing and semantic annotation, applied in neural learning methods, instruments, biological neural network models, etc., and can solve problems such as noise

Active Publication Date: 2020-05-29
BEIHANG UNIV
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] (1) Purpose of the invention: In view of this, the embodiment of the present invention expects to provide a semantic annotation method for hyperspectral remote sensing images, which can at least solve the problem of ignoring the image space context information existing in the existing hyperspectral remote sensing image semantic annotation methods Noise and other technical issues

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
  • A Semantic Annotation Method for Hyperspectral Remote Sensing Images
  • A Semantic Annotation Method for Hyperspectral Remote Sensing Images
  • A Semantic Annotation Method for Hyperspectral Remote Sensing Images

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0082] In order to solve the existing semantic annotation method of hyperspectral image (the flow chart of the existing method is as figure 1 (Shown) existing technical problems such as labeling noise caused by ignoring image spatial context information, embodiments of the present invention provide a semantic labeling method for hyperspectral remote sensing images, which is applied to two or more feature categories Hyperspectral remote sensing images, such as figure 2 Shown.

[0083] The method for semantic annotation of hyperspectral remote sensing images of the present invention has the following steps:

[0084] Step 1: Obtain the training data and test data of the hyperspectral remote sensing image through the spectral information of the hyperspectral remote sensing image and the marked true value.

[0085] The hyperspectral remote sensing image described in this embodiment refers to a remote sensing image that contains dozens or even hundreds of bands of information captured by ...

Embodiment 2

[0150] This embodiment describes the present invention in detail based on an actual scenario.

[0151] The method of this embodiment includes the following steps:

[0152] (1) Generate training data and test data for images.

[0153] The hyperspectral remote sensing image captured by the remote sensing satellite is input into the computer, and the spectral characteristics of the image are normalized. The formula used is as follows:

[0154]

[0155] Where x i Is the i-th training sample point, x is the set of all training sample points, max(x) is the maximum value in the sample matrix, and min(x) is the minimum value in the sample matrix.

[0156] The data set of hyperspectral image generally has only one image. Each image contains a certain number of feature categories, and each feature category contains a different number of sample points. Select and set a sample point from each feature category as the training data, and the remaining sample points as the test data. At the same time...

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 method for semantic labeling of hyperspectral remote sensing images, the steps of which are as follows: 1: Obtain training data and test data of hyperspectral remote sensing images through the spectral information of hyperspectral remote sensing images and label true values; 2: According to the band of hyperspectral remote sensing images Three: Convolutional neural network is trained by training data to obtain the convolutional neural network model; Four: The convolutional neural network model is used to classify the test data to obtain the semantic annotation result; Five: According to the semantic annotation result Construct the unary potential energy function of the conditional random field model; six: Construct the binary potential energy function of the conditional random field model with the edge constraint model based on the improved Mahalanobis distance in the neighborhood; seven: perform the unary potential energy function and binary The weight adjustment of the meta-potential energy function; Eighth: Solving the conditional random field model to obtain the semantic annotation result; through the above steps, a method for semantic annotation of hyperspectral remote sensing images is realized.

Description

Technical field [0001] The invention relates to a semantic annotation method for hyperspectral remote sensing images, and belongs to the technical field of image processing. Background technique [0002] With the rapid development of hyperspectral remote sensing imaging technology, semantic annotation of hyperspectral remote sensing images, as an important means of remote sensing image information extraction, plays an extremely important role in many fields such as disaster monitoring, agricultural surveys, urban planning, and military detection. However, hyperspectral remote sensing images contain huge amount of information, high spectral feature dimensions, complex textures, rich structure and edge details, and imbalance in the number of training samples among different categories. These characteristics have brought great significance to the semantic annotation of hyperspectral remote sensing images. Therefore, the study of efficient and feasible semantic annotation algorithms ...

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): G06K9/62G06N3/08
CPCG06N3/08G06F18/2415
Inventor 姜志国杨俊俐张浩鹏史振威
Owner BEIHANG 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