Analisis Fusi Data Multi Sensor Menggunakan Algoritma Spatial and Temporal Adaptive Reflectance Fusion Model (Studi Kasus : WorldView-3 dan Landsat 8)

Rizkika, Karisma (2018) Analisis Fusi Data Multi Sensor Menggunakan Algoritma Spatial and Temporal Adaptive Reflectance Fusion Model (Studi Kasus : WorldView-3 dan Landsat 8). Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of Undergraduate-Theses_03311440000020.pdf]
Preview
Text
Undergraduate-Theses_03311440000020.pdf - Accepted Version

Download (2MB) | Preview

Abstract

Citra satelit resolusi tinggi cocok digunakan untuk pemetaan tutupan lahan skala besar. Namun karena keterbatasan teknik dan biaya, ketersediaan data multi temporal dapat dikatakan terbatas. Untuk mengatasi masalah tersebut maka dilakukan pengolahan citra, yaitu teknik fusi. Metode fusi yang umum digunakan seperti intensity-hue-saturation (IHS) transformation, principle component substitution (PCS), dan Brovey transformation difokuskan untuk menghasilkan citra dengan spasial dan spektral tinggi pengkombinasian data pankromatik dengan data multispektral yang diperoleh dari citra tersebut secara serempak. Pada penelitian ini, algoritma Spatial and Temporal Adaptive Reflectance Fusion Model (STARFM) dikembangkan untuk mendapatkan data citra dengan resolusi spasial tinggi dan cakupan area secara berkala. Kanal multispektral data multi sensor dari WorldView-3 dan Landsat 8 dimanfaatkan untuk memperoleh hasil fusi yang disebut sebagai citra sintetik. Untuk menguji performa STARFM, uji akurasi dilakukan dengan membuat training sample pada hasil citra sintetik. Akurasi dilakukan dengan melihat keterpisahan antar kelas menggunakan metode matriks konfusi untuk menganalisa akurasi tutupan lahan yang teridentifikasi. Fusi dengan metode ini menunjukkan peningkatan kualitas citra sintetik, diindikasi dengan lebih banyaknya jumlah objek yang dapat diidentifikasi pada citra sintetik (Kappa = 0,68 dan 0,64) yang dikombinasikan dengan citra Landsat asli. Dengan resolusi spasial tinggi dan informasi temporal berkala, maka proses klasifikasi dan interpretasi objek menjadi lebih terbantu guna kebutuhan analisis pemetaan dan pemantauan tutupan lahan.
=====================================================================================================
A high resolution satellite image was suitable for large-scale land cover mapping. However due to technical limitations and budget, the availability of multi-temporal data was limited. Therefore image processing i.e. fusion technique was performed for solving this problem. Common methods such as intensity-hue-saturation (IHS) transformation, principle component substitution (PCS), and Brovey transformation are more focused on generating images which combine high-spatial resolution panchromatic data with multispectral data obtained from the image simultaneously. In this research, multispectral channels from WorldView-3 and Landsat 8 were utilized to obtain fusion result named synthetic images. To assess the performance of STARFM fusion method, an accuracy test was performed by creating sample training based on synthetic images. The accuracy was tested by looking at the separation between classes by the confusion matrix method to analyze the accuracy identified land cover. This fusion was improved the quality of synthetic images that indicated by more identifiable object features on synthetic image (Kappa=0.68 and 0.64) compared to the original Landsat image. With finer spatial feature and more frequent temporal information, the classification and interpretation of object would be useful for land cover mapping and monitoring.

Item Type: Thesis (Undergraduate)
Additional Information: RSG 621.367 8 Riz a-1
Uncontrolled Keywords: fusi citra, citra sintetik, STARFM, resolusi, Image fusion, resolution, synthetic image, STARFM,
Subjects: G Geography. Anthropology. Recreation > G Geography (General) > G70.5.I4 Remote sensing
H Social Sciences > HA Statistics > HA30.6 Spatial analysis
Divisions: Faculty of Civil, Environmental, and Geo Engineering > Geomatics Engineering > 29202-(S1) Undergraduate Theses
Depositing User: Karisma Rizkika
Date Deposited: 27 Jan 2021 03:18
Last Modified: 27 Jan 2021 03:18
URI: http://repository.its.ac.id/id/eprint/53889

Actions (login required)

View Item View Item