Comparing Pixel- and Object-Based Approaches for Classifying Multispectral Drone Imagery of a Salt Marsh Restoration and Reference Site
Monitoring salt marshes with remote sensing is necessary to evaluate their state and restoration. Determining appropriate techniques for this can be overwhelming. Our study provides insight into whether a pixel- or object-based Random Forest classification approach is best for mapping vegetation in...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-03-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/16/6/1049 |
| _version_ | 1797239419462221824 |
|---|---|
| author | Gregory S. Norris Armand LaRocque Brigitte Leblon Myriam A. Barbeau Alan R. Hanson |
| author_facet | Gregory S. Norris Armand LaRocque Brigitte Leblon Myriam A. Barbeau Alan R. Hanson |
| author_sort | Gregory S. Norris |
| collection | DOAJ |
| description | Monitoring salt marshes with remote sensing is necessary to evaluate their state and restoration. Determining appropriate techniques for this can be overwhelming. Our study provides insight into whether a pixel- or object-based Random Forest classification approach is best for mapping vegetation in north temperate salt marshes. We used input variables from drone images (raw reflectances, vegetation indices, and textural features) acquired in June, July, and August 2021 of a salt marsh restoration and reference site in Aulac, New Brunswick, Canada. We also investigated the importance of input variables and whether using landcover classes representing areas of change was a practical way to evaluate variation in the monthly images. Our results indicated that (1) the classifiers achieved overall validation accuracies of 91.1–95.2%; (2) pixel-based classifiers outperformed object-based classifiers by 1.3–2.0%; (3) input variables extracted from the August images were more important than those extracted from the June and July images; (4) certain raw reflectances, vegetation indices, and textural features were among the most important variables; and (5) classes that changed temporally were mapped with user’s and producer’s validation accuracies of 86.7–100.0%. Knowledge gained during this study will inform assessments of salt marsh restoration trajectories spanning multiple years. |
| first_indexed | 2024-04-24T17:51:14Z |
| format | Article |
| id | doaj.art-06c6d6cac2a640e8bc9dc1107fd57291 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-04-24T17:51:14Z |
| publishDate | 2024-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-06c6d6cac2a640e8bc9dc1107fd572912024-03-27T14:02:43ZengMDPI AGRemote Sensing2072-42922024-03-01166104910.3390/rs16061049Comparing Pixel- and Object-Based Approaches for Classifying Multispectral Drone Imagery of a Salt Marsh Restoration and Reference SiteGregory S. Norris0Armand LaRocque1Brigitte Leblon2Myriam A. Barbeau3Alan R. Hanson4Faculty of Forestry and Environmental Management, University of New Brunswick, Fredericton, NB E3B 5A3, CanadaFaculty of Forestry and Environmental Management, University of New Brunswick, Fredericton, NB E3B 5A3, CanadaFaculty of Natural Resource Management, Lakehead University, Thunder Bay, ON P7B 5E1, CanadaDepartment of Biology, University of New Brunswick, Fredericton, NB E3B 5A3, CanadaCanadian Wildlife Service, Environment Canada, P.O. Box 6227, Sackville, NB E4L 4N1, CanadaMonitoring salt marshes with remote sensing is necessary to evaluate their state and restoration. Determining appropriate techniques for this can be overwhelming. Our study provides insight into whether a pixel- or object-based Random Forest classification approach is best for mapping vegetation in north temperate salt marshes. We used input variables from drone images (raw reflectances, vegetation indices, and textural features) acquired in June, July, and August 2021 of a salt marsh restoration and reference site in Aulac, New Brunswick, Canada. We also investigated the importance of input variables and whether using landcover classes representing areas of change was a practical way to evaluate variation in the monthly images. Our results indicated that (1) the classifiers achieved overall validation accuracies of 91.1–95.2%; (2) pixel-based classifiers outperformed object-based classifiers by 1.3–2.0%; (3) input variables extracted from the August images were more important than those extracted from the June and July images; (4) certain raw reflectances, vegetation indices, and textural features were among the most important variables; and (5) classes that changed temporally were mapped with user’s and producer’s validation accuracies of 86.7–100.0%. Knowledge gained during this study will inform assessments of salt marsh restoration trajectories spanning multiple years.https://www.mdpi.com/2072-4292/16/6/1049image classificationecological restorationwetlandpixel-based image analysisobject-based image analysisRandom Forest |
| spellingShingle | Gregory S. Norris Armand LaRocque Brigitte Leblon Myriam A. Barbeau Alan R. Hanson Comparing Pixel- and Object-Based Approaches for Classifying Multispectral Drone Imagery of a Salt Marsh Restoration and Reference Site Remote Sensing image classification ecological restoration wetland pixel-based image analysis object-based image analysis Random Forest |
| title | Comparing Pixel- and Object-Based Approaches for Classifying Multispectral Drone Imagery of a Salt Marsh Restoration and Reference Site |
| title_full | Comparing Pixel- and Object-Based Approaches for Classifying Multispectral Drone Imagery of a Salt Marsh Restoration and Reference Site |
| title_fullStr | Comparing Pixel- and Object-Based Approaches for Classifying Multispectral Drone Imagery of a Salt Marsh Restoration and Reference Site |
| title_full_unstemmed | Comparing Pixel- and Object-Based Approaches for Classifying Multispectral Drone Imagery of a Salt Marsh Restoration and Reference Site |
| title_short | Comparing Pixel- and Object-Based Approaches for Classifying Multispectral Drone Imagery of a Salt Marsh Restoration and Reference Site |
| title_sort | comparing pixel and object based approaches for classifying multispectral drone imagery of a salt marsh restoration and reference site |
| topic | image classification ecological restoration wetland pixel-based image analysis object-based image analysis Random Forest |
| url | https://www.mdpi.com/2072-4292/16/6/1049 |
| work_keys_str_mv | AT gregorysnorris comparingpixelandobjectbasedapproachesforclassifyingmultispectraldroneimageryofasaltmarshrestorationandreferencesite AT armandlarocque comparingpixelandobjectbasedapproachesforclassifyingmultispectraldroneimageryofasaltmarshrestorationandreferencesite AT brigitteleblon comparingpixelandobjectbasedapproachesforclassifyingmultispectraldroneimageryofasaltmarshrestorationandreferencesite AT myriamabarbeau comparingpixelandobjectbasedapproachesforclassifyingmultispectraldroneimageryofasaltmarshrestorationandreferencesite AT alanrhanson comparingpixelandobjectbasedapproachesforclassifyingmultispectraldroneimageryofasaltmarshrestorationandreferencesite |