Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrieval
Learning powerful feature representations for image retrieval has always been a challenging task in the field of remote sensing. Traditional methods focus on extracting low-level hand-crafted features which are not only time-consuming but also tend to achieve unsatisfactory performance due to the co...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2017-05-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | http://www.mdpi.com/2072-4292/9/5/489 |
| _version_ | 1819290949141921792 |
|---|---|
| author | Weixun Zhou Shawn Newsam Congmin Li Zhenfeng Shao |
| author_facet | Weixun Zhou Shawn Newsam Congmin Li Zhenfeng Shao |
| author_sort | Weixun Zhou |
| collection | DOAJ |
| description | Learning powerful feature representations for image retrieval has always been a challenging task in the field of remote sensing. Traditional methods focus on extracting low-level hand-crafted features which are not only time-consuming but also tend to achieve unsatisfactory performance due to the complexity of remote sensing images. In this paper, we investigate how to extract deep feature representations based on convolutional neural networks (CNNs) for high-resolution remote sensing image retrieval (HRRSIR). To this end, several effective schemes are proposed to generate powerful feature representations for HRRSIR. In the first scheme, a CNN pre-trained on a different problem is treated as a feature extractor since there are no sufficiently-sized remote sensing datasets to train a CNN from scratch. In the second scheme, we investigate learning features that are specific to our problem by first fine-tuning the pre-trained CNN on a remote sensing dataset and then proposing a novel CNN architecture based on convolutional layers and a three-layer perceptron. The novel CNN has fewer parameters than the pre-trained and fine-tuned CNNs and can learn low dimensional features from limited labelled images. The schemes are evaluated on several challenging, publicly available datasets. The results indicate that the proposed schemes, particularly the novel CNN, achieve state-of-the-art performance. |
| first_indexed | 2024-12-24T03:30:52Z |
| format | Article |
| id | doaj.art-02376e68b51140838e028f56fcd8e07c |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-12-24T03:30:52Z |
| publishDate | 2017-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-02376e68b51140838e028f56fcd8e07c2022-12-21T17:17:12ZengMDPI AGRemote Sensing2072-42922017-05-019548910.3390/rs9050489rs9050489Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image RetrievalWeixun Zhou0Shawn Newsam1Congmin Li2Zhenfeng Shao3State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, ChinaElectrical Engineering and Computer Science, University of California, Merced, CA 95343, USAState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, ChinaState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, ChinaLearning powerful feature representations for image retrieval has always been a challenging task in the field of remote sensing. Traditional methods focus on extracting low-level hand-crafted features which are not only time-consuming but also tend to achieve unsatisfactory performance due to the complexity of remote sensing images. In this paper, we investigate how to extract deep feature representations based on convolutional neural networks (CNNs) for high-resolution remote sensing image retrieval (HRRSIR). To this end, several effective schemes are proposed to generate powerful feature representations for HRRSIR. In the first scheme, a CNN pre-trained on a different problem is treated as a feature extractor since there are no sufficiently-sized remote sensing datasets to train a CNN from scratch. In the second scheme, we investigate learning features that are specific to our problem by first fine-tuning the pre-trained CNN on a remote sensing dataset and then proposing a novel CNN architecture based on convolutional layers and a three-layer perceptron. The novel CNN has fewer parameters than the pre-trained and fine-tuned CNNs and can learn low dimensional features from limited labelled images. The schemes are evaluated on several challenging, publicly available datasets. The results indicate that the proposed schemes, particularly the novel CNN, achieve state-of-the-art performance.http://www.mdpi.com/2072-4292/9/5/489image retrievaldeep feature representationconvolutional neural networkstransfer learningmulti-layer perceptron |
| spellingShingle | Weixun Zhou Shawn Newsam Congmin Li Zhenfeng Shao Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrieval Remote Sensing image retrieval deep feature representation convolutional neural networks transfer learning multi-layer perceptron |
| title | Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrieval |
| title_full | Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrieval |
| title_fullStr | Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrieval |
| title_full_unstemmed | Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrieval |
| title_short | Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrieval |
| title_sort | learning low dimensional convolutional neural networks for high resolution remote sensing image retrieval |
| topic | image retrieval deep feature representation convolutional neural networks transfer learning multi-layer perceptron |
| url | http://www.mdpi.com/2072-4292/9/5/489 |
| work_keys_str_mv | AT weixunzhou learninglowdimensionalconvolutionalneuralnetworksforhighresolutionremotesensingimageretrieval AT shawnnewsam learninglowdimensionalconvolutionalneuralnetworksforhighresolutionremotesensingimageretrieval AT congminli learninglowdimensionalconvolutionalneuralnetworksforhighresolutionremotesensingimageretrieval AT zhenfengshao learninglowdimensionalconvolutionalneuralnetworksforhighresolutionremotesensingimageretrieval |