Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive Learning
Hyperspectral target detection (HTD) is a crucial aspect of remote sensing applications, aiming to identify targets in hyperspectral images (HSIs) based on their known prior spectral signatures. However, the spectral variability resulting from various imaging conditions in multi-temporal hyperspectr...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/16/4/718 |
| _version_ | 1827342802032263168 |
|---|---|
| author | Xiaodian Zhang Kun Gao Junwei Wang Pengyu Wang Zibo Hu Zhijia Yang Xiaobin Zhao Wei Li |
| author_facet | Xiaodian Zhang Kun Gao Junwei Wang Pengyu Wang Zibo Hu Zhijia Yang Xiaobin Zhao Wei Li |
| author_sort | Xiaodian Zhang |
| collection | DOAJ |
| description | Hyperspectral target detection (HTD) is a crucial aspect of remote sensing applications, aiming to identify targets in hyperspectral images (HSIs) based on their known prior spectral signatures. However, the spectral variability resulting from various imaging conditions in multi-temporal hyperspectral images poses a challenge to both classical and deep learning (DL) methods. To overcome the limitations imposed by spectral variability, an implicit contrastive learning-based target detector (ICLTD) is proposed to exploit in-scene spectra in an unsupervised way. First, only prior spectra are utilized for explicit supervision, while an implicit contrastive learning module (ICLM) is designed to normalize the feature distributions of prior and in-scene spectra. This paper theoretically demonstrates that the ICLM can transfer the gradients from prior spectral features to those of in-scene spectra based on their feature similarities and differences. Because of transferred gradient signals, the ICLTD is regularized to extract similar representations for the prior and in-scene target spectra, while augmenting feature differences between the target and background spectra. Additionally, a local spectral similarity constraint (LSSC) is proposed to enhance the capability of scene adaptation by leveraging the spectral similarities among in-scene targets. To validate the performance of the ICLTD under spectral variability, multi-temporal HSIs captured under various imaging conditions are collected to generate prior spectra and in-scene spectra. Comparative evaluations against several DL detectors and classical methods reveal the superior performance of the ICLTD in achieving a balance between target detectability and background suppressibility under spectral variability. |
| first_indexed | 2024-03-07T22:14:45Z |
| format | Article |
| id | doaj.art-3937e55ff12c41e6a5dcff4b1b73e68c |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-07T22:14:45Z |
| publishDate | 2024-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-3937e55ff12c41e6a5dcff4b1b73e68c2024-02-23T15:33:12ZengMDPI AGRemote Sensing2072-42922024-02-0116471810.3390/rs16040718Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive LearningXiaodian Zhang0Kun Gao1Junwei Wang2Pengyu Wang3Zibo Hu4Zhijia Yang5Xiaobin Zhao6Wei Li7Key Laboratory of Photoelectronic Imaging Technology and System, Beijing Institute of Technology, Beijing 100081, ChinaKey Laboratory of Photoelectronic Imaging Technology and System, Beijing Institute of Technology, Beijing 100081, ChinaKey Laboratory of Photoelectronic Imaging Technology and System, Beijing Institute of Technology, Beijing 100081, ChinaKey Laboratory of Photoelectronic Imaging Technology and System, Beijing Institute of Technology, Beijing 100081, ChinaKey Laboratory of Photoelectronic Imaging Technology and System, Beijing Institute of Technology, Beijing 100081, ChinaKey Laboratory of Photoelectronic Imaging Technology and System, Beijing Institute of Technology, Beijing 100081, ChinaSchool of Information and Electronics, Beijing Institute of Technology, Beijing 100081, ChinaSchool of Information and Electronics, Beijing Institute of Technology, Beijing 100081, ChinaHyperspectral target detection (HTD) is a crucial aspect of remote sensing applications, aiming to identify targets in hyperspectral images (HSIs) based on their known prior spectral signatures. However, the spectral variability resulting from various imaging conditions in multi-temporal hyperspectral images poses a challenge to both classical and deep learning (DL) methods. To overcome the limitations imposed by spectral variability, an implicit contrastive learning-based target detector (ICLTD) is proposed to exploit in-scene spectra in an unsupervised way. First, only prior spectra are utilized for explicit supervision, while an implicit contrastive learning module (ICLM) is designed to normalize the feature distributions of prior and in-scene spectra. This paper theoretically demonstrates that the ICLM can transfer the gradients from prior spectral features to those of in-scene spectra based on their feature similarities and differences. Because of transferred gradient signals, the ICLTD is regularized to extract similar representations for the prior and in-scene target spectra, while augmenting feature differences between the target and background spectra. Additionally, a local spectral similarity constraint (LSSC) is proposed to enhance the capability of scene adaptation by leveraging the spectral similarities among in-scene targets. To validate the performance of the ICLTD under spectral variability, multi-temporal HSIs captured under various imaging conditions are collected to generate prior spectra and in-scene spectra. Comparative evaluations against several DL detectors and classical methods reveal the superior performance of the ICLTD in achieving a balance between target detectability and background suppressibility under spectral variability.https://www.mdpi.com/2072-4292/16/4/718hyperspectral target detectionremote sensingspectral variabilitymulti-temporal hyperspectral images |
| spellingShingle | Xiaodian Zhang Kun Gao Junwei Wang Pengyu Wang Zibo Hu Zhijia Yang Xiaobin Zhao Wei Li Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive Learning Remote Sensing hyperspectral target detection remote sensing spectral variability multi-temporal hyperspectral images |
| title | Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive Learning |
| title_full | Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive Learning |
| title_fullStr | Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive Learning |
| title_full_unstemmed | Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive Learning |
| title_short | Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive Learning |
| title_sort | target detection adapting to spectral variability in multi temporal hyperspectral images using implicit contrastive learning |
| topic | hyperspectral target detection remote sensing spectral variability multi-temporal hyperspectral images |
| url | https://www.mdpi.com/2072-4292/16/4/718 |
| work_keys_str_mv | AT xiaodianzhang targetdetectionadaptingtospectralvariabilityinmultitemporalhyperspectralimagesusingimplicitcontrastivelearning AT kungao targetdetectionadaptingtospectralvariabilityinmultitemporalhyperspectralimagesusingimplicitcontrastivelearning AT junweiwang targetdetectionadaptingtospectralvariabilityinmultitemporalhyperspectralimagesusingimplicitcontrastivelearning AT pengyuwang targetdetectionadaptingtospectralvariabilityinmultitemporalhyperspectralimagesusingimplicitcontrastivelearning AT zibohu targetdetectionadaptingtospectralvariabilityinmultitemporalhyperspectralimagesusingimplicitcontrastivelearning AT zhijiayang targetdetectionadaptingtospectralvariabilityinmultitemporalhyperspectralimagesusingimplicitcontrastivelearning AT xiaobinzhao targetdetectionadaptingtospectralvariabilityinmultitemporalhyperspectralimagesusingimplicitcontrastivelearning AT weili targetdetectionadaptingtospectralvariabilityinmultitemporalhyperspectralimagesusingimplicitcontrastivelearning |