Multiscale Superpixel Guided Discriminative Forest for Hyperspectral Anomaly Detection
Recently, the isolation forest (IF) methods have received increasing attention for their promising performance in hyperspectral anomaly detection (HAD). However, limited by the ability of exploiting spatial-spectral information, existing IF-based methods suffer from a lot of false alarms and disappo...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/19/4828 |
| _version_ | 1827653210999881728 |
|---|---|
| author | Xi Cheng Min Zhang Sheng Lin Kexue Zhou Liang Wang Hai Wang |
| author_facet | Xi Cheng Min Zhang Sheng Lin Kexue Zhou Liang Wang Hai Wang |
| author_sort | Xi Cheng |
| collection | DOAJ |
| description | Recently, the isolation forest (IF) methods have received increasing attention for their promising performance in hyperspectral anomaly detection (HAD). However, limited by the ability of exploiting spatial-spectral information, existing IF-based methods suffer from a lot of false alarms and disappointing performance of detecting local anomalies. To overcome the two problems, a multiscale superpixel guided discriminative forest method is proposed for HAD. First, the multiscale superpixel segmentation is employed to generate some homogeneous regions, and it can effectively extract spatial information to guide anomaly detection for the discriminative forest in local areas. Then, a novel discriminative forest (DF) model with the gain split criterion is designed, which enhances the sensitivity of the DF to local anomalies by the utilization of multi-dimension spectral bands for node division; meanwhile, the acceptable range of hyperplane attribute values is introduced to capture any unseen anomaly pixels that are out-of-range in the evaluation stage. Finally, for the high false alarm rate situation in the existing IF-based algorithms, the multiscale fusion with guided filtering is put forward to refine the initial detection results from the DF. In addition, the extensive experimental results on four real hyperspectral datasets demonstrate the effectiveness of the proposed method. |
| first_indexed | 2024-03-09T21:13:48Z |
| format | Article |
| id | doaj.art-3c41cde5faec4310b82afa65ad4945d6 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T21:13:48Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-3c41cde5faec4310b82afa65ad4945d62023-11-23T21:39:25ZengMDPI AGRemote Sensing2072-42922022-09-011419482810.3390/rs14194828Multiscale Superpixel Guided Discriminative Forest for Hyperspectral Anomaly DetectionXi Cheng0Min Zhang1Sheng Lin2Kexue Zhou3Liang Wang4Hai Wang5School of Aerospace Science and Technology, Xidian University, Xi’an 710126, ChinaSchool of Aerospace Science and Technology, Xidian University, Xi’an 710126, ChinaSchool of Aerospace Science and Technology, Xidian University, Xi’an 710126, ChinaSchool of Aerospace Science and Technology, Xidian University, Xi’an 710126, ChinaShaanxi Academy of Aerospace Technology Application Co., Ltd., Xi’an 710199, ChinaSchool of Aerospace Science and Technology, Xidian University, Xi’an 710126, ChinaRecently, the isolation forest (IF) methods have received increasing attention for their promising performance in hyperspectral anomaly detection (HAD). However, limited by the ability of exploiting spatial-spectral information, existing IF-based methods suffer from a lot of false alarms and disappointing performance of detecting local anomalies. To overcome the two problems, a multiscale superpixel guided discriminative forest method is proposed for HAD. First, the multiscale superpixel segmentation is employed to generate some homogeneous regions, and it can effectively extract spatial information to guide anomaly detection for the discriminative forest in local areas. Then, a novel discriminative forest (DF) model with the gain split criterion is designed, which enhances the sensitivity of the DF to local anomalies by the utilization of multi-dimension spectral bands for node division; meanwhile, the acceptable range of hyperplane attribute values is introduced to capture any unseen anomaly pixels that are out-of-range in the evaluation stage. Finally, for the high false alarm rate situation in the existing IF-based algorithms, the multiscale fusion with guided filtering is put forward to refine the initial detection results from the DF. In addition, the extensive experimental results on four real hyperspectral datasets demonstrate the effectiveness of the proposed method.https://www.mdpi.com/2072-4292/14/19/4828hyperspectral anomaly detectiondiscriminative forestsuperpixel segmentationmultiscale fusionguided filtering |
| spellingShingle | Xi Cheng Min Zhang Sheng Lin Kexue Zhou Liang Wang Hai Wang Multiscale Superpixel Guided Discriminative Forest for Hyperspectral Anomaly Detection Remote Sensing hyperspectral anomaly detection discriminative forest superpixel segmentation multiscale fusion guided filtering |
| title | Multiscale Superpixel Guided Discriminative Forest for Hyperspectral Anomaly Detection |
| title_full | Multiscale Superpixel Guided Discriminative Forest for Hyperspectral Anomaly Detection |
| title_fullStr | Multiscale Superpixel Guided Discriminative Forest for Hyperspectral Anomaly Detection |
| title_full_unstemmed | Multiscale Superpixel Guided Discriminative Forest for Hyperspectral Anomaly Detection |
| title_short | Multiscale Superpixel Guided Discriminative Forest for Hyperspectral Anomaly Detection |
| title_sort | multiscale superpixel guided discriminative forest for hyperspectral anomaly detection |
| topic | hyperspectral anomaly detection discriminative forest superpixel segmentation multiscale fusion guided filtering |
| url | https://www.mdpi.com/2072-4292/14/19/4828 |
| work_keys_str_mv | AT xicheng multiscalesuperpixelguideddiscriminativeforestforhyperspectralanomalydetection AT minzhang multiscalesuperpixelguideddiscriminativeforestforhyperspectralanomalydetection AT shenglin multiscalesuperpixelguideddiscriminativeforestforhyperspectralanomalydetection AT kexuezhou multiscalesuperpixelguideddiscriminativeforestforhyperspectralanomalydetection AT liangwang multiscalesuperpixelguideddiscriminativeforestforhyperspectralanomalydetection AT haiwang multiscalesuperpixelguideddiscriminativeforestforhyperspectralanomalydetection |