Responsive Banner

Deteksi dan klasifikasi tipe bangunan pada Citra Satelit menggunakan Metode K Nearest Neighbor

Jaelani, Adam (2020) Deteksi dan klasifikasi tipe bangunan pada Citra Satelit menggunakan Metode K Nearest Neighbor. Undergraduate thesis, Universitas Islam Negeri Maulana Malik Ibrahim.

[img]
Preview
Text (Fulltext)
14650050.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (7MB) | Preview

Abstract

ABSTRACT

Building density causes various problems in big cities in Indonesia, making smartcities emerge as a solution to reduce the problems that are caused, so that city infrastructure becomes intelligent, the concept of object identification must be used to detect and analyze the presence of cars, roads and buildings. With the presence of satellite imagery, the process of processing digital images associated with objects on the surface of the earth will be very helpful in solving geographical problems, the layout of cities and others. The color segmentation method is used to detect building objects in digital satellite imagery and the k Nearest Neighbormethod is used to classify building objects based on the feature extraction obtained. The average accuracy of building detection with color segmentation is 90.8%, testing the accuracy value of k Nearest Neighbor with k= 5 gets the best accuracy of 93%. The accuracy of the results of the classification of k Nearest Neighborwith k Fold Cross Validationshows the average accuracy of k Nearest Neighbor by 93%.

ABSTRAK

Kepadatan bangunan menimbulkan berbagai permasalahan di kota kota besar di Indonesia, menjadikan kota cerdas muncul menjadi solusi untuk mengurangi masalah yang ditibulkan, agar infrastruk kota menjadi cerdas makas konsep identifikasi objek harus digunakan untuk mendeteksi dan menganalisa keberadaan mobil, jalan, dan bangunan. Dengan adanya citra satelit maka proses pengolahan citra digital terkait dengan objek objek permukaan bumi akan sangat membantu dalam menyelesaikan permasalahan geografis, tataruang kota dan lain lain. Metode segmentasi warna digunakan untuk mendeteksi objek bangunan pada citra satelit digital dan metode k Nearest Neighbor digunakan untuk mengklasifikasi objek bangunan berdasarkan hasil ektraksi ciri yang didapatkan. Rata rata akurasi deteksi bangunan dengan segmentasi warna yaitu 90.8%, pengujian nilai akurasi k Nearest Neighbor dengan k=5 mendapatkan akurasi terbaik 93%. Akurasi hasil pengujian ketepatan klasifikasi k Nearest Neighbor dengan k Fold Cross Validation menujukan rata rata akurasi k Nearest Neighbor sebesar 93%.

Item Type: Thesis (Undergraduate)
Supervisor: Kurniawan, Fachrul and Faisal, Muhammad
Contributors:
ContributionNameEmail
UNSPECIFIEDKurniawan, FachrulUNSPECIFIED
UNSPECIFIEDFaisal, MuhammadUNSPECIFIED
Keywords: Satellite Imagery; Color Segmentation; Building Detection; K Fold Cross Validation; K Nearest Neighbor; Citra Satelit; Segmentasi Warna; Deteksi Bangunan; k Fold Cross Validation; k Nearest Neighbor
Subjects: 08 INFORMATION AND COMPUTING SCIENCES > 0801 Artificial Intelligence and Image Processing > 080106 Image Processing
Departement: Fakultas Sains dan Teknologi > Jurusan Teknik Informatika
Depositing User: Adam Jaelani
Date Deposited: 23 Jul 2020 14:37
Last Modified: 13 Apr 2023 11:32
URI: http://etheses.uin-malang.ac.id/id/eprint/20315

Downloads

Downloads per month over past year

Actions (login required)

View Item View Item