Unified Dynamic Dictionary and Projection Optimization With Full-Rank Representation for Hyperspectral Anomaly Detection
Hyperspectral anomaly detection (HAD) aims to classify each pixel in a hyperspectral image as either background or anomaly without requiring labeled data. Traditional reconstruction based methods model the background using a predefined static background dictionary and low-rank representation coeffic...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10815627/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832582414257356800 |
|---|---|
| author | Hongran Li Chao Wei Yizhou Yang Zhaoman Zhong Ming Xu Dongqing Yuan |
| author_facet | Hongran Li Chao Wei Yizhou Yang Zhaoman Zhong Ming Xu Dongqing Yuan |
| author_sort | Hongran Li |
| collection | DOAJ |
| description | Hyperspectral anomaly detection (HAD) aims to classify each pixel in a hyperspectral image as either background or anomaly without requiring labeled data. Traditional reconstruction based methods model the background using a predefined static background dictionary and low-rank representation coefficients. However, when anomalies are present, the use of a static dictionary can lead to inaccurate background representation, which is easily disturbed by anomalous points. Moreover, existing methods typically focus on the low-rank and smooth characteristics of the background during reconstruction, overlooking deeper features of the background representation. This motivates us to reconsider how the background should be represented. To address these issues, we propose an innovative HAD method that integrates background dictionary learning into the anomaly decomposition process. By using projection operators to optimize the background dictionary, we overcome the limitations of traditional methods that rely on static dictionaries. In addition, we revisit the representation of the background and emphasize the importance of applying nonnegative full-rank constraint to the representation coefficients under the new background dictionary. These improvements result in a more accurate background representation, thereby enhancing anomaly detection performance. Experimental results on several hyperspectral datasets demonstrate that the proposed algorithm excels in anomaly detection tasks, offering new insights and approaches for HAD. |
| format | Article |
| id | doaj-art-eaff4e110b5f4b828ca0da43b9a75a6a |
| institution | Kabale University |
| issn | 1939-1404 2151-1535 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj-art-eaff4e110b5f4b828ca0da43b9a75a6a2025-01-30T00:00:19ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01184032404910.1109/JSTARS.2024.352238810815627Unified Dynamic Dictionary and Projection Optimization With Full-Rank Representation for Hyperspectral Anomaly DetectionHongran Li0https://orcid.org/0000-0002-7437-7359Chao Wei1Yizhou Yang2Zhaoman Zhong3https://orcid.org/0000-0003-0004-3193Ming Xu4Dongqing Yuan5Jiangsu Ocean University, Lianyungang, ChinaJiangsu Ocean University, Lianyungang, ChinaEngineering Research Center of Health Emergency, Jiangsu Provincial Center for Disease Control and Prevention, Nanjing, ChinaJiangsu Ocean University, Lianyungang, ChinaEngineering Research Center of Health Emergency, Jiangsu Provincial Center for Disease Control and Prevention, Nanjing, ChinaJiangsu Ocean University, Lianyungang, ChinaHyperspectral anomaly detection (HAD) aims to classify each pixel in a hyperspectral image as either background or anomaly without requiring labeled data. Traditional reconstruction based methods model the background using a predefined static background dictionary and low-rank representation coefficients. However, when anomalies are present, the use of a static dictionary can lead to inaccurate background representation, which is easily disturbed by anomalous points. Moreover, existing methods typically focus on the low-rank and smooth characteristics of the background during reconstruction, overlooking deeper features of the background representation. This motivates us to reconsider how the background should be represented. To address these issues, we propose an innovative HAD method that integrates background dictionary learning into the anomaly decomposition process. By using projection operators to optimize the background dictionary, we overcome the limitations of traditional methods that rely on static dictionaries. In addition, we revisit the representation of the background and emphasize the importance of applying nonnegative full-rank constraint to the representation coefficients under the new background dictionary. These improvements result in a more accurate background representation, thereby enhancing anomaly detection performance. Experimental results on several hyperspectral datasets demonstrate that the proposed algorithm excels in anomaly detection tasks, offering new insights and approaches for HAD.https://ieeexplore.ieee.org/document/10815627/Background dictionarybackground representation (BR)hyperspectral anomaly detection (HAD)nonnegative full-rank constraint (NFRC)projection operator |
| spellingShingle | Hongran Li Chao Wei Yizhou Yang Zhaoman Zhong Ming Xu Dongqing Yuan Unified Dynamic Dictionary and Projection Optimization With Full-Rank Representation for Hyperspectral Anomaly Detection IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Background dictionary background representation (BR) hyperspectral anomaly detection (HAD) nonnegative full-rank constraint (NFRC) projection operator |
| title | Unified Dynamic Dictionary and Projection Optimization With Full-Rank Representation for Hyperspectral Anomaly Detection |
| title_full | Unified Dynamic Dictionary and Projection Optimization With Full-Rank Representation for Hyperspectral Anomaly Detection |
| title_fullStr | Unified Dynamic Dictionary and Projection Optimization With Full-Rank Representation for Hyperspectral Anomaly Detection |
| title_full_unstemmed | Unified Dynamic Dictionary and Projection Optimization With Full-Rank Representation for Hyperspectral Anomaly Detection |
| title_short | Unified Dynamic Dictionary and Projection Optimization With Full-Rank Representation for Hyperspectral Anomaly Detection |
| title_sort | unified dynamic dictionary and projection optimization with full rank representation for hyperspectral anomaly detection |
| topic | Background dictionary background representation (BR) hyperspectral anomaly detection (HAD) nonnegative full-rank constraint (NFRC) projection operator |
| url | https://ieeexplore.ieee.org/document/10815627/ |
| work_keys_str_mv | AT hongranli unifieddynamicdictionaryandprojectionoptimizationwithfullrankrepresentationforhyperspectralanomalydetection AT chaowei unifieddynamicdictionaryandprojectionoptimizationwithfullrankrepresentationforhyperspectralanomalydetection AT yizhouyang unifieddynamicdictionaryandprojectionoptimizationwithfullrankrepresentationforhyperspectralanomalydetection AT zhaomanzhong unifieddynamicdictionaryandprojectionoptimizationwithfullrankrepresentationforhyperspectralanomalydetection AT mingxu unifieddynamicdictionaryandprojectionoptimizationwithfullrankrepresentationforhyperspectralanomalydetection AT dongqingyuan unifieddynamicdictionaryandprojectionoptimizationwithfullrankrepresentationforhyperspectralanomalydetection |