Segmentasi Ruang Hijau Terbuka Menggunakan Mobile Vision Transformer

Hidayatullah, Arif (2024) Segmentasi Ruang Hijau Terbuka Menggunakan Mobile Vision Transformer. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06111940000107-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (11MB) | Request a copy

Abstract

Ruang Hijau Terbuka (RHT) sangat didukung dan dikembangkan untuk mengurangi permasalahan emisi gas yang tercantum dalam salah satu sustainable development goals. Secara umum, terdapat regulasi dalam penentuan RHT agar suatu wilayah berkontribusi persediaan oksigen dan penyerapan karbondioksida. Untuk menunjang program RHT, perhitungan RHT seharusnya dilakukan sebagai strategi perencanaan. Salah satu perhitungan RHT dapat dijalankan dengan melakukan segmentasi citra dari suatu wilayah. Penelitian Tugas Akhir ini bertujuan untuk segmentasi citra RHT dari suatu wilayah. Penelitian Tugas Akhir ini mengimplementasikan model Mobile Vision Transformer untuk segmentasi citra RHT. Eksperimen dari Tugas Akhir ini menggunakan citra Drone dan Geographic Information System (GIS) dari suatu wilayah. Dari hasil eksperimen, diperoleh bahwa hasil model Mobile Vision Transformer berhasil melakukan segmentasi citra RHT dengan menunjukkan nilai metrik dice coefficient 0.9899 dan intersection over union (IoU) sebesar 0.9801.
=================================================================================================================================
Green Open Space (RHT) is strongly supported and developed to address the issue of gas emissions as stated in one of the sustainable development goals. In general, there are regulations in determining RHT to ensure that an area contributes to oxygen supply and carbon dioxide absorption. To support the RHT program, RHT calculations should be carried out as a planning strategy. One of the RHT calculations can be performed by segmenting images of an area. This study aims to segment RHT images and implement the Mobile Vision Transformer model for the task. This study worked with drone images and Geographic Information System (GIS) images of a specific region. From the experiment results, the Mobile Vision Transformer model successfully segmented RHT images by showing dice coefficient 0.9899 and Jacard Index are 0.9801.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Ruang Hijau Terbuka, Segmentasi Citra, Mobile Vision Transformer
Subjects:	Q Science > QA Mathematics > QA336 Artificial Intelligence
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Arif Hidayatullah
Date Deposited:	12 Feb 2024 08:42
Last Modified:	12 Feb 2024 08:42
URI:	http://repository.its.ac.id/id/eprint/106912

Actions (login required)

View Item