EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Deep Learning Semantic Segmentation for Land Use and Land Cover Types Using Landsat 8 Imagery

Wuttichai Boonpook, Yumin Tan, Attawut Nardkulpat, Kritanai Torsri, Peerapong Torteeka, Patcharin Kamsing, Utane Sawangwit, Jose Pena, Montri Jainaen

ISPRS International Journal of Geo-Information · 07.01.2023 · 12 (1) · 14

DOI: 10.3390/ijgi12010014 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ISPRS International Journal of Geo-Information, online erschienen am 07.01.2023. Verfasst von Wuttichai Boonpook und 8 weiteren Autor:innen. Die Arbeit wurde bisher 73-mal zitiert (Crossref) und nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Faculty of Social Science, Srinakharinwirot University, King Mongkut’s Institute of Technology Ladkrabang, National Science and Technology Development Agency.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Using deep learning semantic segmentation for land use extraction is the most challenging problem in medium spatial resolution imagery. This is because of the deep convolution layer and multiple levels of deep steps of the baseline network, which can cause a degradation problem in small land use features. In this paper, a deep learning semantic segmentation algorithm which comprises an adjustment network architecture (LoopNet) and land use dataset is proposed for automatic land use classification using Landsat 8 imagery. The experimental results illustrate that deep learning semantic segmentation using the baseline network (SegNet, U-Net) outperforms pixel-based machine learning algorithms (MLE, SVM, RF) for land use classification. Furthermore, the LoopNet network, which comprises a convolutional loop and convolutional block, is superior to other baseline networks (SegNet, U-Net, PSPnet) and improvement networks (ResU-Net, DeeplabV3+, U-Net++), with 89.84% overall accuracy and good segmentation results. The evaluation of multispectral bands in the land use dataset demonstrates that Band 5 has good performance in terms of extraction accuracy, with 83.91% overall accuracy. Furthermore, the combination of different spectral bands (Band 1–Band 7) achieved the highest accuracy result (89.84%) compared to individual bands. These results indicate the effectiveness of LoopNet and multispectral bands for land use classification using Landsat 8 imagery.

Förderung

  • Faculty of Social Science, Srinakharinwirot University
  • King Mongkut’s Institute of Technology Ladkrabang
  • National Science and Technology Development Agency

Zitieren

APA

Boonpook, W., Tan, Y., Nardkulpat, A., Torsri, K., Torteeka, P., Kamsing, P., Sawangwit, U., Pena, J., & Jainaen, M. (2023). Deep Learning Semantic Segmentation for Land Use and Land Cover Types Using Landsat 8 Imagery. ISPRS International Journal of Geo-Information, 12(1), 14. https://doi.org/10.3390/ijgi12010014

Vancouver

Boonpook W, Tan Y, Nardkulpat A, Torsri K, Torteeka P, Kamsing P, et al. Deep Learning Semantic Segmentation for Land Use and Land Cover Types Using Landsat 8 Imagery. ISPRS International Journal of Geo-Information. 2023;12(1):14. doi:10.3390/ijgi12010014

BibTeX
@article{Boonpook2023,
  author = {Boonpook, Wuttichai and Tan, Yumin and Nardkulpat, Attawut and Torsri, Kritanai and Torteeka, Peerapong and Kamsing, Patcharin and Sawangwit, Utane and Pena, Jose and Jainaen, Montri},
  title = {Deep Learning Semantic Segmentation for Land Use and Land Cover Types Using Landsat 8 Imagery},
  journal = {ISPRS International Journal of Geo-Information},
  year = {2023},
  volume = {12},
  number = {1},
  pages = {14},
  doi = {10.3390/ijgi12010014}
}
Kostenlos starten →