Data-Wise Spatial Regional Consistency Re-Enhancement for Hyperspectral Image Classification
Effectively using rich spatial and spectral information is the core issue of hyperspectral image (HSI) classification. The recently proposed Diverse Region-based Convolutional Neural Network (DRCNN) achieves good results by weighted averaging the features extracted from several predefined regions, t...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-05-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/9/2227 |
| _version_ | 1797502985578741760 |
|---|---|
| author | Lijian Zhou Erya Xu Siyuan Hao Yuanxin Ye Kun Zhao |
| author_facet | Lijian Zhou Erya Xu Siyuan Hao Yuanxin Ye Kun Zhao |
| author_sort | Lijian Zhou |
| collection | DOAJ |
| description | Effectively using rich spatial and spectral information is the core issue of hyperspectral image (HSI) classification. The recently proposed Diverse Region-based Convolutional Neural Network (DRCNN) achieves good results by weighted averaging the features extracted from several predefined regions, thus exploring the use of spatial consistency to some extent. However, such feature-wise spatial regional consistency enhancement does not effectively address the issue of wrong classifications at the edge of regions, especially when the edge is winding and rough. To improve the feature-wise approach, Data-wise spAtial regioNal Consistency re-Enhancement (“DANCE”) is proposed. Firstly, the HSIs are decomposed once using the Spectral Graph Wavelet (SGW) to enhance the intra-class correlation. Then, the image components in different frequency domains obtained from the weight map are filtered using a Gaussian filter to “debur” the non-smooth region edge. Next, the reconstructed image is obtained based on all filtered frequency domain components using inverse SGW transform. Finally, a DRCNN is used for further feature extraction and classification. Experimental results show that the proposed method achieves the goal of pixel level re-enhancement with image spatial consistency, and can effectively improve not only the performance of the DRCNN, but also that of other feature-wise approaches. |
| first_indexed | 2024-03-10T03:44:01Z |
| format | Article |
| id | doaj.art-8212ba66fe514049b2186d416f76001f |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T03:44:01Z |
| publishDate | 2022-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-8212ba66fe514049b2186d416f76001f2023-11-23T09:12:31ZengMDPI AGRemote Sensing2072-42922022-05-01149222710.3390/rs14092227Data-Wise Spatial Regional Consistency Re-Enhancement for Hyperspectral Image ClassificationLijian Zhou0Erya Xu1Siyuan Hao2Yuanxin Ye3Kun Zhao4School of Information and Control Engineering, Qingdao University of Technology, Qingdao 266525, ChinaSchool of Information and Control Engineering, Qingdao University of Technology, Qingdao 266525, ChinaSchool of Information and Control Engineering, Qingdao University of Technology, Qingdao 266525, ChinaFaculty of Geosciences and Environmental Engineering, Southwest Jiaotong University, Chengdu 610031, ChinaSchool of Information and Control Engineering, Qingdao University of Technology, Qingdao 266525, ChinaEffectively using rich spatial and spectral information is the core issue of hyperspectral image (HSI) classification. The recently proposed Diverse Region-based Convolutional Neural Network (DRCNN) achieves good results by weighted averaging the features extracted from several predefined regions, thus exploring the use of spatial consistency to some extent. However, such feature-wise spatial regional consistency enhancement does not effectively address the issue of wrong classifications at the edge of regions, especially when the edge is winding and rough. To improve the feature-wise approach, Data-wise spAtial regioNal Consistency re-Enhancement (“DANCE”) is proposed. Firstly, the HSIs are decomposed once using the Spectral Graph Wavelet (SGW) to enhance the intra-class correlation. Then, the image components in different frequency domains obtained from the weight map are filtered using a Gaussian filter to “debur” the non-smooth region edge. Next, the reconstructed image is obtained based on all filtered frequency domain components using inverse SGW transform. Finally, a DRCNN is used for further feature extraction and classification. Experimental results show that the proposed method achieves the goal of pixel level re-enhancement with image spatial consistency, and can effectively improve not only the performance of the DRCNN, but also that of other feature-wise approaches.https://www.mdpi.com/2072-4292/14/9/2227hyperspectral image classificationspatial regional consistencySGWTGaussian filtering |
| spellingShingle | Lijian Zhou Erya Xu Siyuan Hao Yuanxin Ye Kun Zhao Data-Wise Spatial Regional Consistency Re-Enhancement for Hyperspectral Image Classification Remote Sensing hyperspectral image classification spatial regional consistency SGWT Gaussian filtering |
| title | Data-Wise Spatial Regional Consistency Re-Enhancement for Hyperspectral Image Classification |
| title_full | Data-Wise Spatial Regional Consistency Re-Enhancement for Hyperspectral Image Classification |
| title_fullStr | Data-Wise Spatial Regional Consistency Re-Enhancement for Hyperspectral Image Classification |
| title_full_unstemmed | Data-Wise Spatial Regional Consistency Re-Enhancement for Hyperspectral Image Classification |
| title_short | Data-Wise Spatial Regional Consistency Re-Enhancement for Hyperspectral Image Classification |
| title_sort | data wise spatial regional consistency re enhancement for hyperspectral image classification |
| topic | hyperspectral image classification spatial regional consistency SGWT Gaussian filtering |
| url | https://www.mdpi.com/2072-4292/14/9/2227 |
| work_keys_str_mv | AT lijianzhou datawisespatialregionalconsistencyreenhancementforhyperspectralimageclassification AT eryaxu datawisespatialregionalconsistencyreenhancementforhyperspectralimageclassification AT siyuanhao datawisespatialregionalconsistencyreenhancementforhyperspectralimageclassification AT yuanxinye datawisespatialregionalconsistencyreenhancementforhyperspectralimageclassification AT kunzhao datawisespatialregionalconsistencyreenhancementforhyperspectralimageclassification |