Automated Built-Up Infrastructure Land Cover Extraction Using Index Ensembles with Machine Learning, Automated Training Data, and Red Band Texture Layers
Automated built-up infrastructure classification is a global need for planning. However, individual indices have weaknesses, including spectral confusion with bare ground, and computational requirements for deep learning are intensive. We present a computationally lightweight method to classify buil...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/16/5/868 |
| _version_ | 1797263941133402112 |
|---|---|
| author | Megan C. Maloney Sarah J. Becker Andrew W. H. Griffin Susan L. Lyon Kristofer Lasko |
| author_facet | Megan C. Maloney Sarah J. Becker Andrew W. H. Griffin Susan L. Lyon Kristofer Lasko |
| author_sort | Megan C. Maloney |
| collection | DOAJ |
| description | Automated built-up infrastructure classification is a global need for planning. However, individual indices have weaknesses, including spectral confusion with bare ground, and computational requirements for deep learning are intensive. We present a computationally lightweight method to classify built-up infrastructure. We use an ensemble of spectral indices and a novel red-band texture layer with global thresholds determined from 12 diverse sites (two seasonally varied images per site). Multiple spectral indexes were evaluated using Sentinel-2 imagery. Our texture metric uses the red band to separate built-up infrastructure from spectrally similar bare ground. Our evaluation produced global thresholds by evaluating ground truth points against a range of site-specific optimal index thresholds across the 24 images. These were used to classify an ensemble, and then spectral indexes, texture, and stratified random sampling guided training data selection. The training data fit a random forest classifier to create final binary maps. Validation found an average overall accuracy of 79.95% (±4%) and an F1 score of 0.5304 (±0.07). The inclusion of the texture metric improved overall accuracy by 14–21%. A comparison to site-specific thresholds and a deep learning-derived layer is provided. This automated built-up infrastructure mapping framework requires only public imagery to support time-sensitive land management workflows. |
| first_indexed | 2024-04-25T00:21:00Z |
| format | Article |
| id | doaj.art-4afb12a1538447c0a40792e5b45437e4 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-04-25T00:21:00Z |
| publishDate | 2024-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-4afb12a1538447c0a40792e5b45437e42024-03-12T16:54:16ZengMDPI AGRemote Sensing2072-42922024-02-0116586810.3390/rs16050868Automated Built-Up Infrastructure Land Cover Extraction Using Index Ensembles with Machine Learning, Automated Training Data, and Red Band Texture LayersMegan C. Maloney0Sarah J. Becker1Andrew W. H. Griffin2Susan L. Lyon3Kristofer Lasko4Geospatial Research Laboratory, Engineering Research and Development Center, US Army Corps of Engineers, Alexandria, VA 22315, USAGeospatial Research Laboratory, Engineering Research and Development Center, US Army Corps of Engineers, Alexandria, VA 22315, USAGeospatial Research Laboratory, Engineering Research and Development Center, US Army Corps of Engineers, Alexandria, VA 22315, USAGeospatial Research Laboratory, Engineering Research and Development Center, US Army Corps of Engineers, Alexandria, VA 22315, USAGeospatial Research Laboratory, Engineering Research and Development Center, US Army Corps of Engineers, Alexandria, VA 22315, USAAutomated built-up infrastructure classification is a global need for planning. However, individual indices have weaknesses, including spectral confusion with bare ground, and computational requirements for deep learning are intensive. We present a computationally lightweight method to classify built-up infrastructure. We use an ensemble of spectral indices and a novel red-band texture layer with global thresholds determined from 12 diverse sites (two seasonally varied images per site). Multiple spectral indexes were evaluated using Sentinel-2 imagery. Our texture metric uses the red band to separate built-up infrastructure from spectrally similar bare ground. Our evaluation produced global thresholds by evaluating ground truth points against a range of site-specific optimal index thresholds across the 24 images. These were used to classify an ensemble, and then spectral indexes, texture, and stratified random sampling guided training data selection. The training data fit a random forest classifier to create final binary maps. Validation found an average overall accuracy of 79.95% (±4%) and an F1 score of 0.5304 (±0.07). The inclusion of the texture metric improved overall accuracy by 14–21%. A comparison to site-specific thresholds and a deep learning-derived layer is provided. This automated built-up infrastructure mapping framework requires only public imagery to support time-sensitive land management workflows.https://www.mdpi.com/2072-4292/16/5/868built-up infrastructure classificationmultispectral classificationtexture metricspectral index thresholds |
| spellingShingle | Megan C. Maloney Sarah J. Becker Andrew W. H. Griffin Susan L. Lyon Kristofer Lasko Automated Built-Up Infrastructure Land Cover Extraction Using Index Ensembles with Machine Learning, Automated Training Data, and Red Band Texture Layers Remote Sensing built-up infrastructure classification multispectral classification texture metric spectral index thresholds |
| title | Automated Built-Up Infrastructure Land Cover Extraction Using Index Ensembles with Machine Learning, Automated Training Data, and Red Band Texture Layers |
| title_full | Automated Built-Up Infrastructure Land Cover Extraction Using Index Ensembles with Machine Learning, Automated Training Data, and Red Band Texture Layers |
| title_fullStr | Automated Built-Up Infrastructure Land Cover Extraction Using Index Ensembles with Machine Learning, Automated Training Data, and Red Band Texture Layers |
| title_full_unstemmed | Automated Built-Up Infrastructure Land Cover Extraction Using Index Ensembles with Machine Learning, Automated Training Data, and Red Band Texture Layers |
| title_short | Automated Built-Up Infrastructure Land Cover Extraction Using Index Ensembles with Machine Learning, Automated Training Data, and Red Band Texture Layers |
| title_sort | automated built up infrastructure land cover extraction using index ensembles with machine learning automated training data and red band texture layers |
| topic | built-up infrastructure classification multispectral classification texture metric spectral index thresholds |
| url | https://www.mdpi.com/2072-4292/16/5/868 |
| work_keys_str_mv | AT megancmaloney automatedbuiltupinfrastructurelandcoverextractionusingindexensembleswithmachinelearningautomatedtrainingdataandredbandtexturelayers AT sarahjbecker automatedbuiltupinfrastructurelandcoverextractionusingindexensembleswithmachinelearningautomatedtrainingdataandredbandtexturelayers AT andrewwhgriffin automatedbuiltupinfrastructurelandcoverextractionusingindexensembleswithmachinelearningautomatedtrainingdataandredbandtexturelayers AT susanllyon automatedbuiltupinfrastructurelandcoverextractionusingindexensembleswithmachinelearningautomatedtrainingdataandredbandtexturelayers AT kristoferlasko automatedbuiltupinfrastructurelandcoverextractionusingindexensembleswithmachinelearningautomatedtrainingdataandredbandtexturelayers |