Wijaya, Aryana Puspa (2026) Pembaruan Tutupan Lahan Peta Rupa Bumi Indonesia Berbasis Analisis Kombinasi Polarisasi Citra SAR (Synthetic Aperture Radar). Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5016221053-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5016221053-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 31 July 2026. Download (8MB) | Request a copy |
Abstract
Peta Rupa Bumi Indonesia (RBI) merupakan peta dasar topografi resmi yang berperan penting dalam menyediakan informasi geospasial sebagai dasar perencanaan dan pengambilan keputusan. Namun seiring dengan terjadinya perubahan tutupan lahan yang berlangsung secara dinamis, menyebabkan informasi yang tersedia pada peta RBI sering kali tidak sesuai dengan kondisi yang aktual di lapangan. Oleh karena itu dibutuhkan metode pembaruan peta RBI yang lebih efisien, salah satunya dengan pemanfaatan penginderaan jauh citra Synthetic Aperture Radar (SAR) Sentinel-1 yang memiliki kelebihan mampu merekam data tanpa dipengaruhi oleh siang dan malam. Penelitian ini bertujuan untuk menganalisis kombinasi polarisasi citra Sentinel-1 dalam klasifikasi tutupan lahan, menilai tingkat akurasi hasil klasifikasi, serta mengidentifikasi faktor-faktor yang mempengaruhi klasifikasi tutupan lahan. Data yang digunakan berupa citra SAR Sentinel-1 dual polarisasi VV dan VH dengan tipe Ground Range Detected (GRD). Lokasi penelitian terletak di Kabupaten Lumajang, Metode penelitian yang dilakukan yaitu dengan tahap pre-processing citra, pembentukan sepuluh skema kombinasi polarisasi, serta klasifikasi tutupan lahan dengan metode unsupervised Expectation Maximazation (EM) Clustering. Kelas tutupan lahan yang diklasifikasikan terdiri dari permukiman, sawah, kebun/perkebunan, tegal/ladang, dan tanah kosong/rumput. Uji Akurasi dilakukan dengan confusion matrix untuk memperoleh nilai overall acurracy dan koefisien kappa. Hasil analisis menunjukkan bahwa kombinasi polarisasi citra Sentinel-1 menghasilkan variasi tampilan citra yang mempengaruhi kemampuan identifikasi tutupan lahan. Permukiman menjadi kelas yang paling mudah dikenali, sedangkan kelas vegetasi dan tanah kosong masih sulit dibedakan. Uji akurasi menunjukkan Skema 1 (VH, VV/VH, VV) dan Skema 6 (VH, VV/VH, VV) sebagai yang terbaik dengan Overall Accuracy sebesar 84% dan nilai Kappa sekitar 0,800. Perbedaan hasil klasifikasi dipengaruhi oleh karakteristik hamburan balik, kondisi permukaan, dan kemiripan antar kelas. Dari segi luas, tidak terdapat satu skema yang paling mendekati data digitasi pada seluruh kelas, di mana Skema 1 (VV, VH, VV/VH) mendekati pada kelas kebun/perkebunan dan ladang, Skema 4 (VV+VH, VH, VV/VH) dan 9 (VV, VH+VV, VH) pada kelas sawah, Skema 7 (VV, VH, VV+VH) pada tanah kosong/rumput, serta Skema 10 (VH+VV, VH, VV) pada permukiman. Secara umum, kombinasi polarisasi Sentinel-1 berpotensi untuk mempercepat dalam identifikasi perubahan tutupan lahan untuk peta RBI Skala 1:25.000.
======================================================================================================================================
The Rupa Bumi Indonesia (RBI) map is an official topographic base map that plays an important role in providing geospatial information as a basis for planning and decision-making. However, along with the occurrence of dynamically changing land cover, the information available on RBI maps often does not correspond to actual conditions in the field. Therefore, a more efficient method for updating RBI maps is needed, one of which is through the utilization of Synthetic Aperture Radar (SAR) remote sensing imagery from Sentinel-1, which has the advantage of being able to record data without being affected by day and night conditions. This study aims to analyze the combination of Sentinel-1 image polarization in land cover classification, assess the accuracy level of the classification results, and identify the factors that influence land cover classification. The data used are Sentinel-1 SAR imagery with dual polarization VV and VH in Ground Range Detected (GRD) type. The study area is located in Lumajang Regency. The research method carried out includes image pre-processing, the formation of ten polarization combination schemes, and land cover classification using the unsupervised Expectation Maximization (EM) Clustering method. The classified land cover classes consist of settlements, rice fields, plantations, dry fields, and bare land/grass. Accuracy testing is conducted using a confusion matrix to obtain the overall accuracy value and the kappa coefficient. The analysis results show that the combination of Sentinel-1 image polarization produces variations in image appearance that affect the ability to identify land cover. Settlements become the easiest class to recognize, while vegetation classes and bare land are still difficult to distinguish. The accuracy test shows Scheme 1 (VH, VV/VH, VV) and Scheme 6 (VH, VV/VH, VV) as the best, with an Overall Accuracy of 84% and a Kappa value of around 0.800. Differences in classification results are influenced by backscatter characteristics, surface conditions, and similarities between classes. In terms of area, there is no single scheme that is closest to the digitized data in all classes, where Scheme 1 (VV, VH, VV/VH) is closest in plantation and dry field classes, Scheme 4 (VV+VH, VH, VV/VH) and Scheme 9 (VV, VH+VV, VH) in rice field classes, Scheme 7 (VV, VH, VV+VH) in bare land/grass, and Scheme 10 (VH+VV, VH, VV) in settlement class. In general, Sentinel-1 polarization combinations have the potential to accelerate the identification of land cover changes for RBI maps at a scale of 1:25,000.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Peta Rupa Bumi Indonesia (RBI), klasifikasi tutupan lahan, Sentinel-1 SAR, kombinasi polarisasi, unsupervised EM, uji akurasi. Indonesian Topographic Base Map (RBI), land cover classification, Sentinel-1 SAR, polarization combination, unsupervised EM, accuracy assessment. |
| Subjects: | T Technology > T Technology (General) > T57.5 Data Processing |
| Divisions: | Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geomatics Engineering > 29202-(S1) Undergraduate Thesis |
| Depositing User: | Aryana Puspa Wijaya |
| Date Deposited: | 24 Jul 2026 06:51 |
| Last Modified: | 24 Jul 2026 06:51 |
| URI: | http://repository.its.ac.id/id/eprint/137462 |
Actions (login required)
 |
View Item |