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

SAR Image Change Detection Method Based on Maximum Margin Metric Learning

A technology of image change detection and metric learning, applied in the field of radar technology remote sensing image processing, can solve problems such as low time complexity, achieve high classification accuracy, improve noise, and improve classification accuracy.

Active Publication Date: 2022-03-18
XIDIAN UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Aiming at the deficiencies of traditional change detection methods and traditional measurement methods, the present invention proposes a SAR image change detection method based on maximum edge metric learning that still maintains high precision with low time complexity

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
  • SAR Image Change Detection Method Based on Maximum Margin Metric Learning
  • SAR Image Change Detection Method Based on Maximum Margin Metric Learning
  • SAR Image Change Detection Method Based on Maximum Margin Metric Learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0045] Compared with optical images, SAR images have their unique advantages: because SAR images are active imaging, SAR has broken through the limitations of optical remote sensing of external conditions such as weather, and has all-weather and all-day working capabilities, and contains phase A variety of information such as , amplitude and polarization make up for the lack of optical images. Therefore, SAR images are widely used, and SAR image change detection is one of its important applications.

[0046] SAR image change detection can be applied to natural disaster assessment. Compared with other imaging, SAR image imaging will not be affected by external conditions such as weather, and images with high imaging quality can still be obtained even under severe conditions. For example, two SAR images before and after the earthquake are obtained, and the degree of disaster after the disaster is observed according to the change detection of the SAR image, which is used to bette...

Embodiment 2

[0058] The SAR image change detection method based on maximum edge metric learning is the same as embodiment 1. In step (4), the positive and negative constraints are given labels, and the specific distribution method is as follows:

[0059] 4.1) If (x 1i ,x 2i )∈S, assign a label y i = 1;

[0060] 4.2) If (x 1i ,x 2i )∈D, assign a label y i =-1.

[0061] The present invention uses 1 and -1 instead of the traditional 1 and 0 when assigning labels, mainly for the positive and negative constraint pairs to work in the optimization process of metric learning. If the negative constraint pair is assigned a label of 0, it will result in The measure of the difference of negative constraints will not affect the optimization of the model, and the distance is always 0.

Embodiment 3

[0063] The SAR image change detection method based on maximum edge metric learning is the same as embodiment 1-2. In step (5), positive and negative constraints are used as input to set up a structured support vector machine model, which specifically includes the following steps:

[0064] The purpose of establishing the metric learning optimization form in the present invention is to find a positive semi-definite matrix A and the corresponding distance threshold b. For the threshold b, after adding a slack variable, for the constraint pair (x 1i ,x 2i )∈S, then the distance between them is less than the threshold b; for the constraint pair (x 1i ,x 2i )∈D, then the distance between them is greater than the threshold b. If the constraint pair S and D are regarded as two categories, then this problem can be transformed into a two-category problem. Because the structured support vector machine model has the minimum generalization error, so the present invention adopts the str...

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 provides a SAR image change detection method based on maximum edge metric learning, which solves the problems that the SAR image change detection is easily affected by coherent speckle noise and traditional metrics cannot measure sample difference information well. The implementation steps are: input remote sensing images before and after the change to construct all samples; construct training samples including all boundaries; use training samples to construct positive and negative constraint pairs; use positive and negative constraint pairs as input, establish a structured support vector machine model to obtain the mapping matrix , decompose the mapping matrix; use the decomposed mapping matrix to map all samples to the feature space, and perform SAR image change detection and classification on all samples in the feature space. The invention has high classification precision, especially maintains high-precision classification effect under the condition of low time complexity, and maintains good boundary information while suppressing noise. For SAR image change detection.

Description

technical field [0001] The invention belongs to the technical field of radar technology remote sensing image processing, and further relates to classification and recognition of remote sensing images, in particular to a SAR image change detection method based on maximum edge metric learning. It is used in disaster assessment, urban development, etc. Background technique [0002] Compared with other imaging methods, the imaging technology of Synthetic Aperture Radar (SAR) has its unique advantages. It can obtain detailed ground object information, and SAR will not be interfered by external factors such as weather conditions and time periods. At the same time, synthetic aperture radar uses synthetic aperture technology to improve its azimuth resolution, and uses pulse compression technology to improve its range resolution, so it can obtain large-area high-resolution remote sensing images, which provides a broad prospect for the application of SAR images. [0003] The existing...

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): G06V20/13G06V10/30G06V10/764G06V10/774G06K9/62
CPCG06V20/13G06V10/30G06F18/214G06F18/243
Inventor 王蓉芳王玉乐陈佳伟焦李成冯婕刘红英尚荣华
Owner XIDIAN 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