Search

Sebagai universitas terkemuka di tanah air, Universitas Gadjah Mada harus siap menjelma sebagai kampus inklusi. Artinya, kampus yang ramah terhadap penyandang diasbilitas. Untuk menjadi kampus inklusi, maka pengelola kampus harus menyediakan fasilitas yang ramah terhadap penyandang disabilitas. Untuk itu penyusunan database spasial bangunan Gedung dengan informasi fasilitas penyandang disabilitas perlu diwujudkan. Basis data yang disusun akan bersumber dari data penginderaan jauh, dan memanfaatkan fasilitas sistem informasi geografis untuk pemanfaatan data spasial. Tujuan penelitian adalah (1) inventarisasi dan menyusun basis data bangunan gedung dan bangunan fasilitas untuk penyandang disabilitas di Kampus Universitas Gadjah Mada, (2) memberikan rekomendasi penyediaan bangunan fasilitas untuk penyandang disabilitas kampus UGM. Interpretasi visual digunakan sebagai metode untuk ekstraksi bangunan Gedung di UGM. Citra penginderaan jauh yang digunakan berupa mosaic foto udara format kecil yang dipotret dengan menggunakan wahana tanpa awak. Survey lapangan dilakukan dengan cara sensus pada seluruh bangunan Gedung di kampus UGM. Penyusunan basis data spasial menggunakan fasilitas sistem informasi geografi. Hasil penelitian berupa data base spasial bangunan Gedung dan bangunan fasilitas untuk penyandang disabilitas di kampus UGM. Atribut yang menyertadi data base tersebut antara lain aksesibilitas bagi penyandang disabilitas berupa ramp atau lift, guiding block, dan fasilitas toilet.

Visual analysis and transformation of vegetation indices have been widely applied in studies of vegetation density using remote sensing data. However, visual analysis is time intensive compared to index transformation. On the other hand, the index transformation from medium resolution imagery is not fully representative for urban vegetation studies. Meanwhile, the spectral range of high-resolution imagery is usually limited to visible wavelengths for the image transformation. Worldview-2 imagery provides a new breakthrough with a high spatial resolution and supports various spectral resolutions. This study aims to explore the spectral value of the Worldview-2 image index for estimation of vegetation density. Normalized indices were made for 56 band combinations and Otsu thresholding was implemented for the threshold selection to separate vegetation and non-vegetation areas. This thresholding was done by minimizing classes’ variances between two groups of pixels which are distinguished by system or classification. The image binarization process was performed to differentiate between vegetation and non-vegetation. For the accuracy testing, a total of 250 samples was produced by a stratified random sampling method. Our results show that the combination of indices from red channel, red-edge, NIR-1, and NIR-2 provides the best accuracy for semantic accuracy. Vegetation area extracted from the index was then compared with the results of the visual analysis. Although the index results in area difference of 2.32 m2 compared to visual analysis, the combination of NIR-2 and red bands can give an accuracy of 96.29 %.

Lahan terbangun di perkotaan dan area vegetasi menjadi hal yang sangat menarik untuk dikaji. Apalagi dinamika penggunaan lahan di perkotaan yang sangat cepat berubah. Berbagai metode dikembangkan untuk ekstraksi lahan terbangun di perkotaan, mulai dari klasifikasi multispektral, object based approach, hingga penelitian berbasis indeks. NDBI menjadi salah satu indeks pioner untuk ekstraksi lahan terbangun perkotaan dengan menggunakan saluran SWIR. Pengembangan indeks lahan terbangun ini masih perlu dikembangan untuk citra yang tidak mempunyai panjang gelombang SWIR. Tujuan penelitian ini adalah merumuskan kombinasi saluran terbaik dalam ekstraksi lahan terbangun dan area vegetasi serta menghitung kepadatan bangunan dan kerapatan vegetasi berbasis indeks. Penelitian ini menggunakan Citra Worldview-2 yang diperoleh dari Digital Globe Foundation untuk ekstraksi lahan terbangun dan kerapatan vegetasi. Normalized difference index digunakan sebagai formula dalam pembuatan indeks. Pemanfaatan semua saluran spektral dalam citra Worldview-2 digunakan untuk ekstraksi lahan terbangun dan kepadatan bangunan di perkotaan dengan PCA sebagai metode untuk penggabungan delapan saluran dalam Worldview-2. Saluran NIR 1 dan NIR 2 yang digabungkan dengan Saluran Merah menjadi pilihan untuk ekstraksi vegetasi. Proses trial dan error mewarnai pemilihan kombinasi saluran yang digunakan dan treshold yang digunakan untuk analisis biner dalam membedakan lahan terbangun dan non lahan terbangun serta area vegetasi dan area non vegetasi. Pemanfaatan unique identification (UID) digunakan untuk pembuatan grid berbasis raster dalam perhitungan kepadatan bangunan dan kerapatan vegetasi. Hasil penelitian menunjukkan bahwa indeks yang dibangun dengan PC2 dan NIR 1 serta PC2 dan NIR 2 mempunyai akurasi tinggi yaitu 94,43% untuk bangunan dan kombinasi indeks dari NIR1_Red mempunyai akurasi optimal yaitu 99,51% dan NIR2_Red mempunyai akurasi 92,87 untuk ekstraksi data vegetasi. Urban phenomenon becomes a very interesting thing to be studied. The urban land use, land conversion, urban green space, are rapidly changing. Various methods were developed for urban built-up data extraction, such as multispectral classification, object-based approach, and index-based research. NDBI became one of pioneer indices for urban-built urban land extraction using SWIR band. The development of this built-up index is still required for images that do not have SWIR wavelengths. The study objectives were to select the best methods for built-up land and vegetation extraction and to calculate building density and index-based vegetation density. Worldview-2 image obtained from Digital Globe Foundation tested for built-up land data extracting and vegetation density analyzing. Normalized difference index formula is applied for combining and setting built-up land and vegetation indexes. Merger of Worldview-2 spectral imagery were using PCA method for extracting built-up land and calculating building density. Combining eight bands into eight new images that have different information from original images was done by PCA method. NIR 1, NIR2, and Red bands are the perfect choice for vegetation extraction because near infrared characteristics have high reflections on vegetation. Selection of band combinations and selection of threshold values through trial and error processes to perceive the best index combinations and reasonable threshold values. Binary analysis is particularly useful for separating the built-up and non-built-up areas as well as vegetation and non-vegetation. The Unique identification (UID) technique used in estimating built-up and vegetation density from precisely classified images provided better and accurate assessment of built-up and vegetation density. The results show that the built-up index involving PC2_NIR 1 and PC2_NIR 2 for the urban built land research achieved an optimal accuracy of 94, 43%. The best accuracy for vegetation data extraction was obtained from the combined NIR1_Red index with 99,51% and NIR2_Red values with an overall accuracy of 92,87%.

Produksi peta penggunaan lahan dari data geospasial satelit harus memperhatikan resolusi spasial yang digunakan, dengan menggunakan ilmu penginderaan jauh secara digital akurasi informasi geospasial tematik yang diperoleh dari data geospasial satelit dapat diukur secara kuantitatif. Jumlah kelas penggunaan lahan, skema klasifikasi, dan teknik ekstraksi informasi akan berpengaruh dalam overall accuracy.Penelitian ini menggunakan tiga variasi jumlah kelas dan skema klasifikasi yaitu 4, 7, dan 10 kelas penggunaan lahan. Tiga variasi tersebut diekstraksi dari citra ALOS AVNIR – 2 dengan resolusi 10 meter menggunakan metode klasifikasi berbasis objek dengan pendekatan support vector machine. Sesuatu yang lain dari klasifikasi berbasis objek adalah proses segmentasi yang mengelompokkan objek tutupan lahan dalam satu bagian, dan algoritma SVM mengklasifikasikan citra segmentasi dengan memanfaatkan empat tipe kernel yaitu linier, radial basis function, sigmoid, dan polynomial. Hasil ekstraksi informasi penggunaan lahan untuk jumlah kelas empat menggunakan klasifikasi berbasis objek dengan pendekatan support vector machine memiliki akurasi sebesar 87.2666% serta nilai koefesien kappa 0.8048 dan tipe kernel yang dipakai adalah kernel Linier. Hasil ekstraksi informasi penggunaan lahan untuk jumlah kelas tujuh menggunakan klasifikasi berbasis objek dengan pendekatan support vector machine memiliki akurasi sebesar 79.8021% serta nilai koefesien kappa 0.7293 dan tipe kernel yang dipakai adalah kernel Linier. Hasil ekstraksi informasi penggunaan lahan untuk jumlah kelas sepuluh menggunakan klasifikasi berbasis objek dengan pendekatan support vector machine memiliki akurasi sebesar 73.3377% serta nilai koefesien kappa 0.6466 dan tipe kernel yang dipakai adalah kernel Linier. Hasil ekstraksi informasi penggunaan lahan untuk jumlah kelas sepuluh menggunakan klasifikasi berbasis objek dengan pendekatan support vector machine dan ditambah dengan data bantu berupa informasi lereng dan informasi pola ruang memiliki akurasi sebesar 77.5705% serta nilai koefesien kappa 0.6982 dan tipe kernel yang dipakai adalah kernel Radial Basis Function. Kernel linier dengan satu parameter menunjukkan akurasi tertinggi pada setiap jumlah kelas yang berbeda. Kernel linier menunjukkkan akurasi terendah ketika dipakai data bantu dalam proses klasifikasi. Berdasarkan hasil penelitian didapat bahwa dengan semakin banyaknya parameter yang dipakai dalam proses klasifikasi tidak berarti semakin akurat hasil klasifikasinya. Setiap kernel memiliki pengaruh spesifik terhadap objek penggunaan lahan, dan untuk meningkatkan akurasi per objek penggunaan lahan dapat menggunakan data bantu

After the 2010 eruption of Merapi, some areas at Sleman Regency experienced with lahars disaster. Many agencies have their own responsibility to overcome the problems in collaborative work. Nevertheless, it is lack of spatial/GIS data support on their decision making as well as on their communications. The main objective of this research was to develop spatial information system to support collaborative lahars disaster management especially for response and recovery phase of the impact on infrastructures at Sleman Regency. Questionnaires and semi-structured interviews were used to investigate their need of the spatial data and the data availability. Basically respondents agreed that spatial data is important in their communication and decision making support. Specifically, the application is utilized with specific spatial analysis tools to support decision making processes. The users’ evaluation of the prototype resulted that almost all of respondents give a good mark to the system.

Impervious surface is one of the major land cover types of urban and suburban environment. Conversion of rural landscapes and vegetation area to urban and suburban land use is directly related to the increase of the impervious surface area. The impervious surface expansion is straight-lined with decreasing green spaces in urban areas. Impervious surface is one of indicator for detecting urban heat islands. This study compares various indices for mapping impervious surfaces using Landsat 8 OLI imagery by optimizing the different spectral characteristics of Landsat 8 OLI imagery. The research objectives are (1) to apply various indices for impervious surface mapping and (2) identifies impervious surfaces in urban areas based on multiple indices and provide recommendations and find the best index for mapping impervious surface in urban areas. In addition to utilizing the index, land use supervised classification method, maximum likelihood classification used for extracting built-up, and non-built-up areas. Accuracy assessment of this research used field data collection as primary data for calculating kappa coefficient, producer accuracy, and user accuracy. The study can also be extended to find the land surface temperature and correlate the impervious surface extraction data with urban heat islands.

Uncontrolled built-up area expansion and building densification could bring some detrimental problems in social and economic aspects such as social inequality, urban heat islands, and disturbance in urban environments. This study monitored multi-decadal building density (1991–2019) in the Yogyakarta urban area, Indonesia consisting of two stages, i.e., built-up area classification and building density estimation, therefore, both built-up expansion and the densification were quantified. Multi sensors of the Landsat series including Landsat 5, 7, and 8 were utilized with some prior corrections to harmonize the reflectance values. A support vector machine (SVM) classifier was used to distinguish between built-up and non built-up areas. Regression algorithms, i.e., linear regression (LR), support vector regression (SVR), and random forest regression (RFR) were explored to obtain the best model to estimate building density using the inputs of built-up indices: Urban Index (UI), Normalized Difference Built-up Index (NDBI), Index-based Built-up Index (IBI), and NIR-based built-up index based on the red (VrNIR-BI) and green band (VgNIR-BI). The best models were revealed by SVR with the inputs of UI-NDBI-IBI and LR with a single predictor of UI, for Landsat 8 (2013–2019) and Landsat 5/7 (1991–2009), respectively, using separate training samples. We found that machine learning regressions (SVM and RF) could perform best when the sample size is abundant, whereas LR could predict better for a limited sample size if a linear positive relationship was identified between the predictor(s) and building density. We conclude that expansion in the study area occurred first, followed by rapid building development in the subsequent years leading to an increase in building density.

A better understanding of the spatial variability of soil nitrogen content is required to achieve the best management in precision agriculture. The purpose of this research was to determine the spatial variability of soil nitrogen content based on reliefs in an apple orchard. The research was conducted from March to August 2018 in a 1210 hectare of apple orchard, Batu, East Java Province, Indonesia. Soil samples were taken using the stratified random sampling method. Data were processed by the GLM Univariate 5% method with SPSS 16.0. The results of the statistical analysis show that the Sig value and coefficient of determination were 0.000 and 0.846, respectively. This condition means that soil nitrogen content was significantly different in various reliefs. The apple orchard was divided into 10 (ten) zones with different soil nitrogen content in various relief. It is crucial as a basis for implementing precision agriculture in apple orchards, meaning that the determination of fertilizer dosage is adjusted to the soil nitrogen content in the various zones. This study concluded that relief significantly affects the spatial variability of soil nitrogen content.