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

Heterogeneous remote sensing image change detection method and device based on deep learning

A remote sensing image and change detection technology, which is applied in the field of image processing, can solve problems such as low precision and misjudgment of changed areas, and achieve the effect of solving data differences and improving visual effects

Pending Publication Date: 2021-04-09
ELECTRIC POWER RES INST OF STATE GRID ZHEJIANG ELECTRIC POWER COMAPNY +4
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, most of these methods are based on unsupervised ideas, and finally use the threshold segmentation method to process the results, which can easily lead to misjudgment of changing areas, so the accuracy is often not high
[0008] In summary, the current heterogeneous image change detection methods have certain defects, and an image change detection method with high completion and accuracy is urgently needed

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
  • Heterogeneous remote sensing image change detection method and device based on deep learning
  • Heterogeneous remote sensing image change detection method and device based on deep learning
  • Heterogeneous remote sensing image change detection method and device based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0058] Embodiment 1 provides a method for detecting changes in heterogeneous remote sensing images, aiming at detecting changes in heterogeneous remote sensing images through the combination of GAN and change detection network, which has high precision and is universal in the field of remote sensing detection.

[0059] Please refer to figure 1 As shown, a heterogeneous remote sensing image change detection method includes the following steps:

[0060] S110. Receive multi-temporal heterogeneous remote sensing images;

[0061] The multi-temporal heterogeneous remote sensing images received in S110, in order to increase the accuracy of subsequent model identification and analysis, also need to be preprocessed, including geometric correction, atmospheric correction, and georeferencing; geometric correction is mainly to prevent image and reference system The deformation generated when the expression requirements in are inconsistent. Atmospheric correction is mainly to eliminate th...

Embodiment 2

[0077] The second embodiment mainly explains and illustrates the training process of the preset GAN network and the preset change detection network in the first embodiment.

[0078] Please refer to image 3 as shown,

[0079] The training process of the preset GAN network comprises the following steps:

[0080] S210. Receive two remote sensing images of X source and Y source;

[0081] S220. Construct the data sets of the remote sensing images of the X source and the Y source respectively, and divide the data sets into a training set and a test set;

[0082] When dividing the data set in S220, the remote sensing images can be cropped with a 50% overlapping degree, and each image can be cropped into m images of n×n size, n≥256; wherein, in this embodiment, the remote sensing images are required to be color RGB 3-channel image or grayscale image.

[0083] The proportion of the training set and the testing set in S220 is not specifically limited in this embodiment, and can be ...

Embodiment 3

[0114] Embodiment 3 discloses a device corresponding to the heterogeneous remote sensing image change detection method corresponding to the above embodiment, which is the virtual device structure of the above embodiment, please refer to Figure 6 shown, including:

[0115] A receiving module 410, configured to receive multi-temporal heterogeneous remote sensing images;

[0116] A conversion module 420, configured to input the heterogeneous remote sensing image into a preset GAN network for image source conversion to obtain a single-source remote sensing image;

[0117] The output module 430 is configured to input the single-source remote sensing image into a preset change detection network to obtain a binary change map.

[0118] Preferably, receiving multi-temporal heterogeneous remote sensing images includes the following steps:

[0119] Preprocessing is performed on the multi-temporal heterogeneous remote sensing image, and the preprocessing includes geometric correction, ...

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 heterologous remote sensing image change detection method based on deep learning, relates to the technical field of image processing, and is used for solving the problems of inaccurate change detection and single image source, and the method comprises the following steps: receiving multi-time heterologous remote sensing images; inputting the heterologous remote sensing image into a preset GAN network for image source conversion to obtain a single-source remote sensing image; and inputting the single-source remote sensing image into a preset change detection network to obtain a binary change graph. The invention also discloses a heterologous remote sensing image change detection device which is used for converting the heterologous remote sensing image and obtaining a binary change graph of the remote sensing image through the change detection network. According to the invention, domain conversion of two remote sensing images can be realized at the same time; the problem of data difference between different remote sensing images can be effectively solved, the change areas of different time phase images are extracted through the deep learning change detection network, and the visual effect is improved to a certain extent.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method and device for detecting changes in heterogeneous remote sensing images based on deep learning. Background technique [0002] When processing remote sensing images, change detection is one of the very important topics. Change detection plays an important role in practical applications such as disaster relief, agricultural survey, urban planning, and military monitoring. With the advancement of remote sensing image processing technology, the emergence of various resolutions and various sensors has increased the diversity of remote sensing data, providing sufficient data protection for change detection. When performing change detection, it is usually necessary to detect change trends, and the detection often only relies on one data form, resulting in certain limitations in terms of response time to emergencies and time resolution. [0003] Compared with only rely...

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): G06K9/00G06K9/62G06T7/246G06N3/04G06N3/08
CPCG06T7/246G06N3/08G06T2207/10032G06T2207/20081G06T2207/20084G06V20/13G06N3/048G06N3/045G06F18/214
Inventor 袁兆祥杜正舜李星华司为国韩嘉佳汪自翔罗少杰刘剑吴发献方炯徐晓华
Owner ELECTRIC POWER RES INST OF STATE GRID ZHEJIANG ELECTRIC POWER COMAPNY
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