High-efficiency scene classification based on deep compressed-domain feature
Remote sensing image (RSI) scene classification has become a more and more fundamental issue in satellite and UAV time-sensitive applications. However, as the volume and velocity of RSIs are undergoing an explosive growth, traditional effective technologies claim a huge amount of computing resources...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2019-08-01
|
Series: | The Journal of Engineering |
Subjects: | |
Online Access: | https://digital-library.theiet.org/content/journals/10.1049/joe.2019.0266 |
id |
doaj-d2a6a6bdf8b64de49d71983227b68a74 |
---|---|
record_format |
Article |
spelling |
doaj-d2a6a6bdf8b64de49d71983227b68a742021-04-02T12:26:22ZengWileyThe Journal of Engineering2051-33052019-08-0110.1049/joe.2019.0266JOE.2019.0266High-efficiency scene classification based on deep compressed-domain featureCheng Li0Baojun Zhao1Boya Zhao2Wenzheng Wang3Chenhui Duan4Beijing Institute of TechnologyBeijing Institute of TechnologyBeijing Institute of TechnologyBeijing Institute of TechnologyBeijing Institute of TechnologyRemote sensing image (RSI) scene classification has become a more and more fundamental issue in satellite and UAV time-sensitive applications. However, as the volume and velocity of RSIs are undergoing an explosive growth, traditional effective technologies claim a huge amount of computing resources, on a scale that already goes beyond in-orbit processing capacity. Here, we attempt to design a deep feature representation framework based on onboard compressed data to solve the aforementioned problem. Firstly, we extract header and body information in raw JPEG2000 codestream for representing geometrical and geospatial property of RSIs. Moreover, a novel compressed-domain convolutional neural network (CNN) is proposed to obtain high-level representation for effective classification. Extensive experiments demonstrate that, in comparison with the existing relevant state-of-the-art approaches, the proposed method achieves high classification accuracy with faster computation and lower consumption.https://digital-library.theiet.org/content/journals/10.1049/joe.2019.0266remote sensingimage codingdata compressiongeophysical image processingimage classificationfeature extractionimage representationconvolutional neural netscomputing resourcesin-orbit processing capacitydeep feature representation frameworkbody informationraw JPEG2000 codestreamcompressed-domain convolutional neural networkhigh-level representationeffective classificationhigh classification accuracyhigh-efficiency scene classificationcompressed-domain featureremote sensing image scene classificationRSIsatellite applicationheader informationUAV time-sensitive application |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Cheng Li Baojun Zhao Boya Zhao Wenzheng Wang Chenhui Duan |
spellingShingle |
Cheng Li Baojun Zhao Boya Zhao Wenzheng Wang Chenhui Duan High-efficiency scene classification based on deep compressed-domain feature The Journal of Engineering remote sensing image coding data compression geophysical image processing image classification feature extraction image representation convolutional neural nets computing resources in-orbit processing capacity deep feature representation framework body information raw JPEG2000 codestream compressed-domain convolutional neural network high-level representation effective classification high classification accuracy high-efficiency scene classification compressed-domain feature remote sensing image scene classification RSI satellite application header information UAV time-sensitive application |
author_facet |
Cheng Li Baojun Zhao Boya Zhao Wenzheng Wang Chenhui Duan |
author_sort |
Cheng Li |
title |
High-efficiency scene classification based on deep compressed-domain feature |
title_short |
High-efficiency scene classification based on deep compressed-domain feature |
title_full |
High-efficiency scene classification based on deep compressed-domain feature |
title_fullStr |
High-efficiency scene classification based on deep compressed-domain feature |
title_full_unstemmed |
High-efficiency scene classification based on deep compressed-domain feature |
title_sort |
high-efficiency scene classification based on deep compressed-domain feature |
publisher |
Wiley |
series |
The Journal of Engineering |
issn |
2051-3305 |
publishDate |
2019-08-01 |
description |
Remote sensing image (RSI) scene classification has become a more and more fundamental issue in satellite and UAV time-sensitive applications. However, as the volume and velocity of RSIs are undergoing an explosive growth, traditional effective technologies claim a huge amount of computing resources, on a scale that already goes beyond in-orbit processing capacity. Here, we attempt to design a deep feature representation framework based on onboard compressed data to solve the aforementioned problem. Firstly, we extract header and body information in raw JPEG2000 codestream for representing geometrical and geospatial property of RSIs. Moreover, a novel compressed-domain convolutional neural network (CNN) is proposed to obtain high-level representation for effective classification. Extensive experiments demonstrate that, in comparison with the existing relevant state-of-the-art approaches, the proposed method achieves high classification accuracy with faster computation and lower consumption. |
topic |
remote sensing image coding data compression geophysical image processing image classification feature extraction image representation convolutional neural nets computing resources in-orbit processing capacity deep feature representation framework body information raw JPEG2000 codestream compressed-domain convolutional neural network high-level representation effective classification high classification accuracy high-efficiency scene classification compressed-domain feature remote sensing image scene classification RSI satellite application header information UAV time-sensitive application |
url |
https://digital-library.theiet.org/content/journals/10.1049/joe.2019.0266 |
work_keys_str_mv |
AT chengli highefficiencysceneclassificationbasedondeepcompresseddomainfeature AT baojunzhao highefficiencysceneclassificationbasedondeepcompresseddomainfeature AT boyazhao highefficiencysceneclassificationbasedondeepcompresseddomainfeature AT wenzhengwang highefficiencysceneclassificationbasedondeepcompresseddomainfeature AT chenhuiduan highefficiencysceneclassificationbasedondeepcompresseddomainfeature |
_version_ |
1721569031306084352 |