Spatial Prior Fuzziness Pool-Based Interactive Classification of Hyperspectral Images
Acquisition of labeled data for supervised Hyperspectral Image (HSI) classification is expensive in terms of both time and costs. Moreover, manual selection and labeling are often subjective and tend to induce redundancy into the classifier. Active learning (AL) can be a suitable approach for HSI cl...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-05-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/11/9/1136 |
| _version_ | 1798017742513635328 |
|---|---|
| author | Muhammad Ahmad Asad Khan Adil Mehmood Khan Manuel Mazzara Salvatore Distefano Ahmed Sohaib Omar Nibouche |
| author_facet | Muhammad Ahmad Asad Khan Adil Mehmood Khan Manuel Mazzara Salvatore Distefano Ahmed Sohaib Omar Nibouche |
| author_sort | Muhammad Ahmad |
| collection | DOAJ |
| description | Acquisition of labeled data for supervised Hyperspectral Image (HSI) classification is expensive in terms of both time and costs. Moreover, manual selection and labeling are often subjective and tend to induce redundancy into the classifier. Active learning (AL) can be a suitable approach for HSI classification as it integrates data acquisition to the classifier design by ranking the unlabeled data to provide advice for the next query that has the highest training utility. However, multiclass AL techniques tend to include redundant samples into the classifier to some extent. This paper addresses such a problem by introducing an AL pipeline which preserves the most representative and spatially heterogeneous samples. The adopted strategy for sample selection utilizes fuzziness to assess the mapping between actual output and the approximated a-posteriori probabilities, computed by a marginal probability distribution based on discriminative random fields. The samples selected in each iteration are then provided to the spectral angle mapper-based objective function to reduce the inter-class redundancy. Experiments on five HSI benchmark datasets confirmed that the proposed Fuzziness and Spectral Angle Mapper (FSAM)-AL pipeline presents competitive results compared to the state-of-the-art sample selection techniques, leading to lower computational requirements. |
| first_indexed | 2024-04-11T16:12:15Z |
| format | Article |
| id | doaj.art-c85139428eba49fa9b2a243f66bd644e |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-04-11T16:12:15Z |
| publishDate | 2019-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-c85139428eba49fa9b2a243f66bd644e2022-12-22T04:14:40ZengMDPI AGRemote Sensing2072-42922019-05-01119113610.3390/rs11091136rs11091136Spatial Prior Fuzziness Pool-Based Interactive Classification of Hyperspectral ImagesMuhammad Ahmad0Asad Khan1Adil Mehmood Khan2Manuel Mazzara3Salvatore Distefano4Ahmed Sohaib5Omar Nibouche6Dipartimento di Matematica e Informatica—MIFT, University of Messina, Messina 98121, ItalySchool of Computer Science, South China Normal University, Guangzhou 510000, ChinaInstitute of Data Science and Artificial Intelligence, Innopolis University, Innopolis 420500, RussiaInstitute of Software Development and Engineering, Innopolis University, Innopolis 420500, RussiaDipartimento di Matematica e Informatica—MIFT, University of Messina, Messina 98121, ItalyDepartment of Computer Engineering, Khwaja Fareed University of Engineering and Information Technology, Rahim Yar Khan 64200, PakistanFaculty of Computing, Engineering and the Built Environment, Ulster University, Newtownabbey, Co Antrim BT37 0QB, UKAcquisition of labeled data for supervised Hyperspectral Image (HSI) classification is expensive in terms of both time and costs. Moreover, manual selection and labeling are often subjective and tend to induce redundancy into the classifier. Active learning (AL) can be a suitable approach for HSI classification as it integrates data acquisition to the classifier design by ranking the unlabeled data to provide advice for the next query that has the highest training utility. However, multiclass AL techniques tend to include redundant samples into the classifier to some extent. This paper addresses such a problem by introducing an AL pipeline which preserves the most representative and spatially heterogeneous samples. The adopted strategy for sample selection utilizes fuzziness to assess the mapping between actual output and the approximated a-posteriori probabilities, computed by a marginal probability distribution based on discriminative random fields. The samples selected in each iteration are then provided to the spectral angle mapper-based objective function to reduce the inter-class redundancy. Experiments on five HSI benchmark datasets confirmed that the proposed Fuzziness and Spectral Angle Mapper (FSAM)-AL pipeline presents competitive results compared to the state-of-the-art sample selection techniques, leading to lower computational requirements.https://www.mdpi.com/2072-4292/11/9/1136hyperspectral imagingactive learningfuzzinessspectral angle mappersoft threshold |
| spellingShingle | Muhammad Ahmad Asad Khan Adil Mehmood Khan Manuel Mazzara Salvatore Distefano Ahmed Sohaib Omar Nibouche Spatial Prior Fuzziness Pool-Based Interactive Classification of Hyperspectral Images Remote Sensing hyperspectral imaging active learning fuzziness spectral angle mapper soft threshold |
| title | Spatial Prior Fuzziness Pool-Based Interactive Classification of Hyperspectral Images |
| title_full | Spatial Prior Fuzziness Pool-Based Interactive Classification of Hyperspectral Images |
| title_fullStr | Spatial Prior Fuzziness Pool-Based Interactive Classification of Hyperspectral Images |
| title_full_unstemmed | Spatial Prior Fuzziness Pool-Based Interactive Classification of Hyperspectral Images |
| title_short | Spatial Prior Fuzziness Pool-Based Interactive Classification of Hyperspectral Images |
| title_sort | spatial prior fuzziness pool based interactive classification of hyperspectral images |
| topic | hyperspectral imaging active learning fuzziness spectral angle mapper soft threshold |
| url | https://www.mdpi.com/2072-4292/11/9/1136 |
| work_keys_str_mv | AT muhammadahmad spatialpriorfuzzinesspoolbasedinteractiveclassificationofhyperspectralimages AT asadkhan spatialpriorfuzzinesspoolbasedinteractiveclassificationofhyperspectralimages AT adilmehmoodkhan spatialpriorfuzzinesspoolbasedinteractiveclassificationofhyperspectralimages AT manuelmazzara spatialpriorfuzzinesspoolbasedinteractiveclassificationofhyperspectralimages AT salvatoredistefano spatialpriorfuzzinesspoolbasedinteractiveclassificationofhyperspectralimages AT ahmedsohaib spatialpriorfuzzinesspoolbasedinteractiveclassificationofhyperspectralimages AT omarnibouche spatialpriorfuzzinesspoolbasedinteractiveclassificationofhyperspectralimages |