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

A method for hyperspectral image classification based on separable 3D residual networks and transfer learning

A hyperspectral image and transfer learning technology, which is applied in biological neural network models, character and pattern recognition, instruments, etc., can solve the problems of shallow deep learning models, and achieve the effect of deep network models, high precision, and fewer parameters

Active Publication Date: 2022-05-10
NORTHWESTERN POLYTECHNICAL UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, limited by the number of hyperspectral image training samples, the deep learning model applied to hyperspectral image classification is relatively shallow, although a large number of experiments in computer vision have shown that effectively increasing the depth is very beneficial for improving classification performance

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 method for hyperspectral image classification based on separable 3D residual networks and transfer learning
  • A method for hyperspectral image classification based on separable 3D residual networks and transfer learning
  • A method for hyperspectral image classification based on separable 3D residual networks and transfer learning

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment

[0039] Step 1: Data preprocessing. The hyperspectral image data to be processed is subjected to maximum and minimum normalization, and the normalization formula is as follows:

[0040]

[0041]Step 2: Data partitioning. Data partitioning. For the pre-training dataset, all labeled samples are extracted as the pre-training dataset. For the target data set, 10-20 samples of each category are taken as the training set, and the rest are used as the test set. The specific method of extracting samples is as follows. For a three-dimensional hyperspectral image data with a size of M×N×L, M and N represent the height and width of the hyperspectral image respectively, and L represents the number of bands of the data. When extracting samples, take the pixel to be processed as the center, and extract a data block of S×S×L as the sample data of the central pixel, and S represents the size of the neighborhood, generally 27.

[0042] Step 3: Build a network model. The network designed...

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 present invention involves a three -dimensional convolutional network based on a separate three -dimensional residual network and migration learning high spectral image classification. First of all, the design parameters are less and suitable for high spectrum image characteristics.Secondly, the migration technology between high spectrum images obtained by different sensors is combined with the three -dimensional convolutional network model to achieve high -precision classification of high -spectrum images under small sample conditions.Under the conditions of small sample conditions, the depth of the depth of high spectrometer image, the classification of high accuracy.Compared with the existing high -spectrum image classification method based on deep learning based on deep learning, the network model is deeper, more accurate, and less parameters.

Description

technical field [0001] The invention relates to a hyperspectral image classification method based on a separable three-dimensional residual network and transfer learning, which belongs to the field of remote sensing image processing. Background technique [0002] Hyperspectral images contain both spectral information and spatial information, and have important applications in military and civilian fields. However, the high-dimensional characteristics of hyperspectral images, high correlation between bands, and spectral mixing make hyperspectral image classification face great challenges. In recent years, with the emergence of new technologies of deep learning, the hyperspectral image classification method based on deep learning has made a breakthrough. However, deep learning models usually contain a large number of parameters and require a large number of training samples. However, there are relatively few annotated samples of hyperspectral images, and it is difficult to f...

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): G06V10/764G06V10/82G06K9/62G06N3/04
Inventor 李映姜晔楠张号逵呼延烺
Owner NORTHWESTERN POLYTECHNICAL 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