Satellite Image Matching Method Based on Deep Convolutional Neural Network
This article focuses on the first aspect of the album of deep learning: the deep convolutional method. The traditional matching point extraction algorithm typically uses manually designed feature descriptors and the shortest distance between them to match as the matching criterion. The matching resu...
Main Author: | |
---|---|
Format: | Article |
Language: | English |
Published: |
Surveying and Mapping Press
2019-06-01
|
Series: | Journal of Geodesy and Geoinformation Science |
Subjects: | |
Online Access: | http://jggs.sinomaps.com/fileup/2096-5990/PDF/1584692322090-726962763.pdf |
id |
doaj-a53c60c53e704db29678da1709fbea03 |
---|---|
record_format |
Article |
spelling |
doaj-a53c60c53e704db29678da1709fbea032020-11-25T02:53:20ZengSurveying and Mapping PressJournal of Geodesy and Geoinformation Science2096-59902019-06-01229010010.11947/j.JGGS.2019.0210Satellite Image Matching Method Based on Deep Convolutional Neural NetworkDazhao FAN,Yang DONG,Yongsheng ZHANG0Institute of Geospatial Information, Information Engineering University, Zhengzhou 450001, ChinaThis article focuses on the first aspect of the album of deep learning: the deep convolutional method. The traditional matching point extraction algorithm typically uses manually designed feature descriptors and the shortest distance between them to match as the matching criterion. The matching result can easily fall into a local extreme value, which causes missing of the partial matching point. Targeting this problem, we introduce a two-channel deep convolutional neural network based on spatial scale convolution, which performs matching pattern learning between images to realize satellite image matching based on a deep convolutional neural network. The experimental results show that the method can extract the richer matching points in the case of heterogeneous, multi-temporal and multi-resolution satellite images, compared with the traditional matching method. In addition, the accuracy of the final matching results can be maintained at above 90%.http://jggs.sinomaps.com/fileup/2096-5990/PDF/1584692322090-726962763.pdf|image matching|deep learning|convolutional neural network|satellite image |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Dazhao FAN,Yang DONG,Yongsheng ZHANG |
spellingShingle |
Dazhao FAN,Yang DONG,Yongsheng ZHANG Satellite Image Matching Method Based on Deep Convolutional Neural Network Journal of Geodesy and Geoinformation Science |image matching|deep learning|convolutional neural network|satellite image |
author_facet |
Dazhao FAN,Yang DONG,Yongsheng ZHANG |
author_sort |
Dazhao FAN,Yang DONG,Yongsheng ZHANG |
title |
Satellite Image Matching Method Based on Deep Convolutional Neural Network |
title_short |
Satellite Image Matching Method Based on Deep Convolutional Neural Network |
title_full |
Satellite Image Matching Method Based on Deep Convolutional Neural Network |
title_fullStr |
Satellite Image Matching Method Based on Deep Convolutional Neural Network |
title_full_unstemmed |
Satellite Image Matching Method Based on Deep Convolutional Neural Network |
title_sort |
satellite image matching method based on deep convolutional neural network |
publisher |
Surveying and Mapping Press |
series |
Journal of Geodesy and Geoinformation Science |
issn |
2096-5990 |
publishDate |
2019-06-01 |
description |
This article focuses on the first aspect of the album of deep learning: the deep convolutional method. The traditional matching point extraction algorithm typically uses manually designed feature descriptors and the shortest distance between them to match as the matching criterion. The matching result can easily fall into a local extreme value, which causes missing of the partial matching point. Targeting this problem, we introduce a two-channel deep convolutional neural network based on spatial scale convolution, which performs matching pattern learning between images to realize satellite image matching based on a deep convolutional neural network. The experimental results show that the method can extract the richer matching points in the case of heterogeneous, multi-temporal and multi-resolution satellite images, compared with the traditional matching method. In addition, the accuracy of the final matching results can be maintained at above 90%. |
topic |
|image matching|deep learning|convolutional neural network|satellite image |
url |
http://jggs.sinomaps.com/fileup/2096-5990/PDF/1584692322090-726962763.pdf |
work_keys_str_mv |
AT dazhaofanyangdongyongshengzhang satelliteimagematchingmethodbasedondeepconvolutionalneuralnetwork |
_version_ |
1724725125624889344 |