Use of a DNN-Based Image Translator with Edge Enhancement Technique to Estimate Correspondence between SAR and Optical Images
In this paper, the local correspondence between synthetic aperture radar (SAR) images and optical images is proposed using an image feature-based keypoint-matching algorithm. To achieve accurate matching, common image features were obtained at the corresponding locations. Since the appearance of SAR...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-04-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/12/9/4159 |
| _version_ | 1797505793723990016 |
|---|---|
| author | Hisatoshi Toriya Ashraf Dewan Hajime Ikeda Narihiro Owada Mahdi Saadat Fumiaki Inagaki Youhei Kawamura Itaru Kitahara |
| author_facet | Hisatoshi Toriya Ashraf Dewan Hajime Ikeda Narihiro Owada Mahdi Saadat Fumiaki Inagaki Youhei Kawamura Itaru Kitahara |
| author_sort | Hisatoshi Toriya |
| collection | DOAJ |
| description | In this paper, the local correspondence between synthetic aperture radar (SAR) images and optical images is proposed using an image feature-based keypoint-matching algorithm. To achieve accurate matching, common image features were obtained at the corresponding locations. Since the appearance of SAR and optical images is different, it was difficult to find similar features to account for geometric corrections. In this work, an image translator, which was built with a DNN (deep neural network) and trained by conditional generative adversarial networks (cGANs) with edge enhancement, was employed to find the corresponding locations between SAR and optical images. When using conventional cGANs, many blurs appear in the translated images and they degrade keypoint-matching accuracy. Therefore, a novel method applying an edge enhancement filter in the cGANs structure was proposed to find the corresponding points between SAR and optical images to accurately register images from different sensors. The results suggested that the proposed method could accurately estimate the corresponding points between SAR and optical images. |
| first_indexed | 2024-03-10T04:23:25Z |
| format | Article |
| id | doaj.art-fb82977032f048bdbb63ca8aecd8f097 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T04:23:25Z |
| publishDate | 2022-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-fb82977032f048bdbb63ca8aecd8f0972023-11-23T07:45:04ZengMDPI AGApplied Sciences2076-34172022-04-01129415910.3390/app12094159Use of a DNN-Based Image Translator with Edge Enhancement Technique to Estimate Correspondence between SAR and Optical ImagesHisatoshi Toriya0Ashraf Dewan1Hajime Ikeda2Narihiro Owada3Mahdi Saadat4Fumiaki Inagaki5Youhei Kawamura6Itaru Kitahara7Faculty of International Resource Sciences, Akita University, 1-1 Tegatagakuen-Machi, Akita-City 0100862, Akita, JapanSchool of Earth and Planetary Sciences, Curtin University, Kent St. Bentley, WA 6102, AustraliaFaculty of International Resource Sciences, Akita University, 1-1 Tegatagakuen-Machi, Akita-City 0100862, Akita, JapanFaculty of International Resource Sciences, Akita University, 1-1 Tegatagakuen-Machi, Akita-City 0100862, Akita, JapanFaculty of International Resource Sciences, Akita University, 1-1 Tegatagakuen-Machi, Akita-City 0100862, Akita, JapanFaculty of International Resource Sciences, Akita University, 1-1 Tegatagakuen-Machi, Akita-City 0100862, Akita, JapanFaculty of Engineering, Hokkaido University, Kita 8, Nishi 5, Kita-Ku, Sapporo-City 0608628, Hokkaido, JapanCenter for Computational Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba-City 3058577, Ibaraki, JapanIn this paper, the local correspondence between synthetic aperture radar (SAR) images and optical images is proposed using an image feature-based keypoint-matching algorithm. To achieve accurate matching, common image features were obtained at the corresponding locations. Since the appearance of SAR and optical images is different, it was difficult to find similar features to account for geometric corrections. In this work, an image translator, which was built with a DNN (deep neural network) and trained by conditional generative adversarial networks (cGANs) with edge enhancement, was employed to find the corresponding locations between SAR and optical images. When using conventional cGANs, many blurs appear in the translated images and they degrade keypoint-matching accuracy. Therefore, a novel method applying an edge enhancement filter in the cGANs structure was proposed to find the corresponding points between SAR and optical images to accurately register images from different sensors. The results suggested that the proposed method could accurately estimate the corresponding points between SAR and optical images.https://www.mdpi.com/2076-3417/12/9/4159image registrationkeypoint matchingsynthetic aperture radardeep neural networkgenerative adversarial networks |
| spellingShingle | Hisatoshi Toriya Ashraf Dewan Hajime Ikeda Narihiro Owada Mahdi Saadat Fumiaki Inagaki Youhei Kawamura Itaru Kitahara Use of a DNN-Based Image Translator with Edge Enhancement Technique to Estimate Correspondence between SAR and Optical Images Applied Sciences image registration keypoint matching synthetic aperture radar deep neural network generative adversarial networks |
| title | Use of a DNN-Based Image Translator with Edge Enhancement Technique to Estimate Correspondence between SAR and Optical Images |
| title_full | Use of a DNN-Based Image Translator with Edge Enhancement Technique to Estimate Correspondence between SAR and Optical Images |
| title_fullStr | Use of a DNN-Based Image Translator with Edge Enhancement Technique to Estimate Correspondence between SAR and Optical Images |
| title_full_unstemmed | Use of a DNN-Based Image Translator with Edge Enhancement Technique to Estimate Correspondence between SAR and Optical Images |
| title_short | Use of a DNN-Based Image Translator with Edge Enhancement Technique to Estimate Correspondence between SAR and Optical Images |
| title_sort | use of a dnn based image translator with edge enhancement technique to estimate correspondence between sar and optical images |
| topic | image registration keypoint matching synthetic aperture radar deep neural network generative adversarial networks |
| url | https://www.mdpi.com/2076-3417/12/9/4159 |
| work_keys_str_mv | AT hisatoshitoriya useofadnnbasedimagetranslatorwithedgeenhancementtechniquetoestimatecorrespondencebetweensarandopticalimages AT ashrafdewan useofadnnbasedimagetranslatorwithedgeenhancementtechniquetoestimatecorrespondencebetweensarandopticalimages AT hajimeikeda useofadnnbasedimagetranslatorwithedgeenhancementtechniquetoestimatecorrespondencebetweensarandopticalimages AT narihiroowada useofadnnbasedimagetranslatorwithedgeenhancementtechniquetoestimatecorrespondencebetweensarandopticalimages AT mahdisaadat useofadnnbasedimagetranslatorwithedgeenhancementtechniquetoestimatecorrespondencebetweensarandopticalimages AT fumiakiinagaki useofadnnbasedimagetranslatorwithedgeenhancementtechniquetoestimatecorrespondencebetweensarandopticalimages AT youheikawamura useofadnnbasedimagetranslatorwithedgeenhancementtechniquetoestimatecorrespondencebetweensarandopticalimages AT itarukitahara useofadnnbasedimagetranslatorwithedgeenhancementtechniquetoestimatecorrespondencebetweensarandopticalimages |