DOI: https://doi.org/10.14710/elipsoida.2025.25145

EKSTRAKSI OTOMATIS TAPAK BANGUNAN (BUILDING FOOTPRINT) PADA ORTOFOTO MENGGUNAKAN SEGMENT ANYTHING MODEL (SAM)

1Master of Geomatics Engineering, Departement of Geodesy, Faculty of Engineering, Universitas Gadjah Mada, Indonesia

2Departmenent of Geodetic Engineering, Indonesia

3Faculty of Engineering, Indonesia

4 Universitas Gadjah Mada, Indonesia

View all affiliations
Received: 22 Mar 2025; Accepted: 19 Jun 2025; Published: 19 Jun 2025.

Citation Format:
Abstract
Kebutuhan peta dasar skala besar khususnya di Indonesia semakin meningkat seiring berkembangnya waktu. Salah satu fitur geospasial penting adalah bangunan. Penggunaan teknologi UAV mampu menghasilkan produk ortofoto resolusi tinggi yang dapat membantu pemetaan skala besar. Kumpulan data bangunan yang diturunkan dari data ortofoto dapat memberikan informasi untuk pemantauan dan perencanaan suatu daerah, khususnya di daerah yang tidak memiliki peta perencanaan rinci atau data kadaster. Pada umumnya, ekstraksi objek tapak bangunan dilakukan dengan digitasi manual. Namun, perkembangan terbaru menunjukkan teknologi otomatisasi dengan deep learning memiliki keunggulan lebih baik dari sisi kinerja yang lebih singkat. Teknik deep learning yang terbaru saat ini adalah SAM (Segment Anything Model). SAM merupakan pendekatan baru yang dikembangkan oleh Meta AI untuk segmentasi yang telah dilatih pada dataset sangat besar sehingga tidak memerlukan pelatihan ulang (Kirillov et al., 2023). Penelitian ini memanfaatkan SAM untuk ekstraksi tapak bangunan khususnya wilayah Indonesia. Karakteristik bangunan yang sangat bervariasi menjadi tantangan algoritma SAM dalam mengekstraksi tapak bangunan. Selain SAM, penggunaan metode regularisasi diterapkan untuk memperbaiki bentuk bangunan hasil segmentasi yang tidak teratur dan tegas. Hasil uji akurasi precision, recall, f1-score, dan IoU secara keseluruhan menunjukkan rata-rata nilainya diatas 87 %. Hasil tersebut menunjukkan bahwa SAM mampu melakukan ekstraksi tapak bangunan dengan baik.
Fulltext View|Download
Funding: Universitas Gajah Mada

Article Metrics:

Article Info
Section: Articles
Language : ID
Recent articles
EKSTRAKSI OTOMATIS TAPAK BANGUNAN (BUILDING FOOTPRINT) PADA ORTOFOTO MENGGUNAKAN SEGMENT ANYTHING MODEL (SAM) Scanning the Lawang Sewu for Eternal Historical 3D Documentation using Terrestrial Laser Scanner INTEGRASI CITRA SENTINEL-1 DAN SENTINEL-2 UNTUK PEMETAAN TUTUPAN LAHAN TAHUN 2024 (STUDI KASUS: BWP I IKN DAN SEKITARNYA) More recent articles
  1. Abdollahi, A., Pradhan, B., Gite, S., & Alamri, A. (2020). Building Footprint Extraction from High Resolution Aerial Images Using Generative Adversarial Network (GAN) Architecture. IEEE Access, 8, 209517–209527. https://doi.org/10.1109/ACCESS.2020.3038225
  2. Elachi, C., & Van Zyl, J. (2021). Introduction to the Physics and Techniques of Remote Sensing, Third Edition. In A JOHN WILEY & SONS, INC., PUBLICATION. https://doi.org/10.1002/9781119523048
  3. Farajzadeh, Z., Saadatseresht, M., Alidoost, F., Engineering, G., & Science, C. (2023). AUTOMATIC BUILDING EXTRACTION FROM UAV-BASED IMAGES AND DSMs USING DEEP LEARNING. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, X-4/W1-202(February), 171–177
  4. Gui, B., Bhardwaj, A., & Sam, L. (2024). Evaluating the Efficacy of Segment Anything Model for Delineating Agriculture and Urban Green Spaces in Multiresolution Aerial and Spaceborne Remote Sensing Images. Remote Sensing, 16(2), 1–32. https://doi.org/10.3390/rs16020414
  5. Guo, H., Su, X., Tang, S., Du, B., & Zhang, L. (2021). Scale-Robust Deep-Supervision Network for Mapping Building Footprints from High-Resolution Remote Sensing Images. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 14, 10091–10100. https://doi.org/10.1109/JSTARS.2021.3109237
  6. Janiesch, C., Zschech, P., & Heinrich, K. (2023). Machine learning and deep learning. Electronics Markets, 31, 71–84. https://doi.org/10.1515/9783110791402-004
  7. Ji, W., Li, J., Bi, Q., Liu, T., Li, W., & Cheng, L. (2023). Segment Anything Is Not Always Perfect: An Investigation of SAM on Different Real-world Applications. http://arxiv.org/abs/2304.05750
  8. Jiang, S., Jiang, C., & Jiang, W. (2020). Efficient structure from motion for large-scale UAV images: A review and a comparison of SfM tools. ISPRS Journal of Photogrammetry and Remote Sensing, 167(September 2019), 230–251. https://doi.org/10.1016/j.isprsjprs.2020.04.016
  9. Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A. C., Lo, W.-Y., Dollár, P., & Girshick, R. (2023). Segment Anything. Arxiv Computer Vision and Pattern Recognition, 1–30. http://arxiv.org/abs/2304.02643
  10. Li, Y., Wang, D., Yuan, C., Li, H., & Hu, J. (2023). Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model Adapter. Sensors, 23(18). https://doi.org/10.3390/s23187884
  11. Negara, T. B., & Harintaka. (2021). Pemodelan Bangunan 3D Menggunakan Footprint Bangunan Hasil Ekstraksi Mask R-CNN dan Dense Point Cloud dari Foto Udara UAV. Prosiding FIT ISI, 1, 248–260
  12. Osco, L. P., Wu, Q., de Lemos, E. L., Gonçalves, W. N., Ramos, A. P. M., Li, J., & Marcato, J. (2023). The Segment Anything Model (SAM) for remote sensing applications: From zero to one shot. International Journal of Applied Earth Observation and Geoinformation, 124(October), 1–18. https://doi.org/10.1016/j.jag.2023.103540
  13. Rastogi, K., Bodani, P., & Sharma, S. A. (2022). Automatic building footprint extraction from very high-resolution imagery using deep learning techniques. Geocarto International, 37(5), 1501–1513. https://doi.org/10.1080/10106049.2020.1778100
  14. Ren, Y., Yang, X., Wang, Z., Yu, G., Liu, Y., Liu, X., Meng, D., Zhang, Q., & Yu, G. (2023). Segment Anything Model (SAM) Assisted Remote Sensing Supervision for Mariculture—Using Liaoning Province, China as an Example. Remote Sensing, 15(24). https://doi.org/10.3390/rs15245781
  15. Rezatofighi, H., Tsoi, N., Gwak, J., Sadeghian, A., Reid, I., & Savarese, S. (2019). Generalized intersection over union: A metric and a loss for bounding box regression. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2019-June, 658–666. https://doi.org/10.1109/CVPR.2019.00075
  16. Wu, Q., & Osco, L. P. (2023). samgeo: A Python package for segmenting geospatial data with the Segment Anything Model (SAM). Journal of Open Source Software, 8(89), 5663. https://doi.org/10.21105/joss.05663
  17. Yang, H., Wu, P., Yao, X., Wu, Y., Wang, B., & Xu, Y. (2018). Building extraction in very high resolution imagery by dense-attention networks. Remote Sensing, 10(11), 1–16. https://doi.org/10.3390/rs10111768
  18. Yao, H., Qin, R., & Chen, X. (2019). Unmanned aerial vehicle for remote sensing applications - A review. Remote Sensing, 11(12), 1–22. https://doi.org/10.3390/rs11121443
  19. Yu, T., Tang, P., Zhao, B., Bai, S., Gou, P., Liao, J., & Jin, C. (2023). ConvBNet: A Convolutional Network for Building Footprint Extraction. IEEE Geoscience and Remote Sensing Letters, 20, 1–5. https://doi.org/10.1109/LGRS.2023.3250091

Last update:

No citation recorded.

Last update:

No citation recorded.