Pemodelan Bangunan LoD2 dari Data LiDAR UAV Menggunakan Algoritma Sifter (Spatial Intelligent Footprint Extraction And Reconstruction)

Hamdi, Wildan (2026) Pemodelan Bangunan LoD2 dari Data LiDAR UAV Menggunakan Algoritma Sifter (Spatial Intelligent Footprint Extraction And Reconstruction). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5016221006-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (18MB) | Request a copy

Abstract

Perkembangan teknologi geospasial mendorong kebutuhan akan data tiga dimensi (3D) yang mampu merepresentasikan objek bangunan secara lebih rinci. Salah satu bentuk representasi tersebut adalah model bangunan tingkat Level of Detail 2 (LoD2), yang menampilkan informasi geometri bangunan berupa tinggi dan bentuk atap. Penelitian ini bertujuan menyusun model bangunan 3D LoD2 dari data LiDAR UAV dengan memanfaatkan building Footprint berbasis pendekatan Algoritma SIFTER (Spatial Intelligent Footprint Extraction and Reconstruction) pada wilayah Kecamatan Waru, Kabupaten Sidoarjo. Data yang digunakan meliputi hasil pengukuran GNSS, Orthophoto UAV, data LiDAR UAV, serta data Footprint bangunan awal dan atribut Morfometri. Tahapan penelitian mencakup pengolahan data GNSS, pembentukan Orthophoto, DTM, DSM, dan nDSM, klasifikasi Footprint bangunan berbasis Python menggunakan model XGBoost, seleksi kandidat bangunan LoD2, ekstraksi atribut tinggi bangunan, penyiapan atribut atap dan warna, serta pemodelan prosedural menggunakan Computer Generated Architecture (CGA) di ArcGIS CityEngine. Hasil penelitian menunjukkan bahwa model klasifikasi menghasilkan tingkat kesesuaian prediksi terhadap label sebesar 0,9978 (99,78%). Tahap seleksi kandidat menghasilkan 208 bangunan siap model dari 239 objek yang dievaluasi. Ekstraksi atribut tinggi bangunan menghasilkan rata-rata tinggi 6,295 meter, median 5,756 meter, dan tinggi maksimum 10,970 meter. Evaluasi model dilakukan melalui penilaian kualitas geometrik data sumber dan evaluasi visual terhadap Orthophoto UAV, DSM, dan nDSM. Hasil evaluasi menunjukkan bahwa model yang dihasilkan memiliki kesesuaian yang baik terhadap bentuk Footprint, tinggi relatif bangunan, bentuk atap, dan tampilan visual bangunan.
==================================================================================================================================
The development of geospatial technology has increased the need for three-dimensional (3D) data capable of representing buildings in greater detail. One such representation is a building model at Level of Detail 2 (LoD2), which includes geometric information such as building height and roof shape. This study aims to develop a 3D LoD2 building model from UAV LiDAR data by utilizing a building Footprint based on the SIFTER algorithm approach (Spatial Intelligent Footprint Extraction and Reconstruction) in Waru District, Sidoarjo Regency. The data used include GNSS measurements, UAV Orthophotos, UAV LiDAR data, and initial building Footprint data with morphometric attributes. The research stages consist of GNSS data processing, Orthophoto generation, DTM, DSM, and nDSM production, Python-based building Footprint classification using the XGBoost model, selection of LoD2 building candidates, extraction of building height attributes, preparation of roof and color attributes, and procedural modeling using Computer Generated Architecture (CGA) in ArcGIS CityEngine. The results show that the classification model achieved a prediction-to-label agreement of 0.9978 (99,78%). The candidate selection stage produced 208 model-ready buildings out of 239 evaluated objects. Building height extraction resulted in an average height of 6.295 meters, a median of 5.756 meters, and a maximum height of 10.970 meters. Model evaluation was carried out through assessment of source data geometric quality and visual evaluation against UAV Orthophotos, DSM, and nDSM. The evaluation results indicate that the generated model has good agreement in terms of Footprint shape, relative building height, roof shape, and visual appearance.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	ArcGIS CityEngine, LiDAR UAV, LoD2, Orthophoto UAV, SIFTER, XGBoost
Subjects:	G Geography. Anthropology. Recreation > G Geography (General) G Geography. Anthropology. Recreation > G Geography (General) > G109.5 Global Positioning System G Geography. Anthropology. Recreation > G Geography (General) > G70.212 ArcGIS. Geographic information systems. G Geography. Anthropology. Recreation > G Geography (General) > G70.217 Geospatial data T Technology > TA Engineering (General). Civil engineering (General) > TA593 Orthophotography
Divisions:	Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geomatics Engineering > 29202-(S1) Undergraduate Thesis
Depositing User:	Wildan Hamdi
Date Deposited:	22 Jul 2026 07:04
Last Modified:	22 Jul 2026 07:04
URI:	http://repository.its.ac.id/id/eprint/136130

Actions (login required)

View Item