Analisis Algoritma Random Forest dan Kombinasi Indeks Spektral untuk Identifikasi Lahan Terbangun (Kasus Kota Surakarta)

Ardia Tiara Rahmi(1*), Kholis Hapsari Pratiwi(2), Delista Putri Deni(3),

(1) Universitas Sebelas Maret, Jawa Tengah, Indonesia
(2) Universitas Sebelas Maret, Jawa Tengah, Indonesia
(3) Universitas Sebelas Maret, Jawa Tengah, Indonesia
(*) Corresponding Author

Abstract


In an effort to realize one of the objectives of the Sustainable Development Goals (SDGs) program goal 11 concerning Sustainable Cities and Settlements, controlling the intensity of urban built-up land in Indonesia really needs attention. One effort to monitor the condition of changes in built-up land that is easy, fast, cheap and efficient is to use the results of remote sensing data processing. Technological developments in the field of remote sensing are currently leading to processing based on Big Data and cloud computing, one of which is GEE (Google Earth Engine). Google Earth Engine. Identification of built-up land was carried out using the guided classification method of Machine Learning Random Forest and a combination of Spectral Index algorithms consisting of a combination of Urban Index (UI), Normalized Difference Vegetation Index (NDVI), and Modified Normalized Difference Water Index (MNDWI) algorithms. The object segmentation process on the results of the combined spectral index is carried out using the Otsu thresholding method. The results showed that the built-up area from the results of identification using the Random Forest algorithm and the spectral index combination algorithm in 2019 to 2023 has increased the area of land, which is shown in the 2019-2023 built-up land map. The accuracy test using the confusion matrix showed that the results of identification using a combination of spectral indices obtained OA and Kappa values which were included in the medium to high category, namely 98.69% and 0.969. So that this method can then continue to be used to monitor the pattern of development of built-up land in Surakarta City.

Full Text:

PDF

References


Puspitasari, S., & Suharyadi. (2016). Kajian Kepadatan Bangunan Menggunakan Interpretasi Hibrida Citra Landsat-8 OLI di Kota Semarang Tahun 2015. Jurnal Bumi Indonesia, 5, 1–9.

BPS Pemerintah Kota Surabaya. (2016). Surakarta dalam angka 2022. Badan Pusat Statistik Kota Surakarta Tahun 2022.

Shelestov, Andrii, Mykola Lavreniuk, Nataliia Kussul, and Alexei Novikov. 2017. “Exploring Google Earth Engine Platform for Big Data Processing : Classification of Multi-Temporal Satellite Imagery for Crop Mapping.” 5(February): 1–10. https:// doi.org/10.3389/feart.2017.00017

Mutanga, Onisimo, and Lalit Kumar. 2019. “Google Earth Engine Applications.” Remote Sensing 11(591): 11–14. https://doi.org/10.3390/ rs11050591

Teluguntla, P., & Thenkabail, P. (2018). A 30-m landsat-derived cropland extent product of Australia and China using random forest machine learning algorithm on Google Earth Engine cloud computing platform. ISPRS J. Photogramm. Remote Sens., 325-340

Kurniawan, S., Nurhaidar, W. O., & Salihin, I. (2017). Optimalisasi Transformasi Spektral UI, NDBI, NDVI dan Kombinasi Transformasi Spektral UI-NDVI dan NDBI-NDVI Guna Mendeteksi Kepadatan Lahan Terbangun di Kota Magelang. Jurnal Geografi Aplikasi Dan Teknologi, 1 (1), 13–22.

Hidayati, I. N., Suharyadi, R., & Danoedoro, P. (2018b). Kombinasi Indeks Citra untuk Analisis Lahan Terbangun dan Vegetasi Perkotaan. Majalah Geografi Indonesia, 32 (1), 24. https://doi.org/10.22146/mgi.31899

Mustofa, Dian (2018). Perbandingan Metode Klasifikasi Berbasis Machine Learning Pada Google Earth Engine Untuk Pemetaan Perubahan Penutup Lahan (Studi Kasus: Daerah Aliran Sungai Opak-Oyo). Diunduh dari http://etd.repository.ugm.ac.id/.

Kamal, M., Farda, N. M., Jamaluddin, I., Parela, A., Wikantika, K., Prasetyo, L. B., & Irawan, B. (2020). A preliminary study on machine learning and google earth engine for mangrove mapping. IOP Conference Series: Earth and Environmental Science (Vol. 500, No. 1, p. 012038). IOP Publishing.




DOI: http://dx.doi.org/10.30645/j-sakti.v7i2.692

Refbacks

  • There are currently no refbacks.



J-SAKTI (Jurnal Sains Komputer & Informatika)
Published Papers Indexed/Abstracted By:


Jumlah Kunjungan :

View My Stats