View Item 
      •   IPB Repository
      • Dissertations and Theses
      • Undergraduate Theses
      • UT - Faculty of Forestry and Environment
      • UT - Forest Management
      • View Item
      •   IPB Repository
      • Dissertations and Theses
      • Undergraduate Theses
      • UT - Faculty of Forestry and Environment
      • UT - Forest Management
      • View Item
      JavaScript is disabled for your browser. Some features of this site may not work without it.

      Kajian Metode Mesin Vektor Pendukung dan Peluang Maksimum pada Citra Pleiades-1B Berbasis Segmentasi Algoritma Mean-Shift: Studi Kasus di Hutan Lahan Kering Kalimantan Utara

      Thumbnail
      View/Open
      Cover (455.4Kb)
      Full text (1.343Mb)
      Lampiran (1.338Mb)
      Date
      2023-02-01
      Author
      Amarullah, Nadian
      Jaya, I Nengah Surati
      Metadata
      Show full item record
      Abstract
      Studi ini meneliti tentang algoritma segmentasi mean-shift menggunakan pendekatan klasifikasi parametrik yaitu peluang maksimum (MLC) dan pendekatan klasifikasi non-parametrik mesin vektor pendukung (SVM) pada citra satelit resolusi sangat tinggi Pleiades-1B dalam mendeteksi penutupan tajuk ekosistem hutan lahan kering. Pendekatan klasifikasi MLC dan SVM digunakan untuk mengklasifikasi hasil segmentasi mean-shift menjadi tajuk dan bukan tajuk. Klasifikasi segmentasi citra Pleiades-1B menggunakanMLC menghasilkan akurasi kappa (KA) tertinggi sebesar 50% pada kombinasi K-08, sedangkan menggunakan SVM menghasilkan KA sebesar 73% pada kombinasi K-07. Akurasi tertinggi hasil segmentasi didapat menggunakan parameter range radius (Hr) senilai 20 dan akurasi terendah menggunakan Hr senilai 18. Hasil studi menunjukkan bahwa hasil terbaik klasifikasi segmentasi mean-shift yaitu menggunakan klasifikasi nonparametrik SVM dengan parameter segmentasi Hr senilai 20. Kata Kunci: akurasi kappa, mean-shift, MLC, SVM, penutupan tajuk.
       
      This study examined the mean-shift segmentation algorithm using the parametric classification approach of Maximum likelihood (MLC) and the non-parametric classification approach of Support Vector Machine (SVM) on very high-resolution satellite images of Pleiades-1B in detecting canopy closure of dryland forest ecosystems. The MLC and SVM classification approaches were used to classify the results of mean-shift segmentation into canopy and non-canopy. The segmented Pleiades-1B image classification using the MLC provided the highest KA of 50% in combination K-08, while using SVM resulted a KA of 72% with the combination of K- 07. The highest accuracy of segmentation classification results was obtained using a range radius (Hr) parameter of 20, while the lowest accuracy was with Hr of 18. The study results show that the best result of mean-shift segmentation classification is using SVM non-parametric classification with a segmentation parameter Hr of 20.
       
      URI
      http://repository.ipb.ac.id/handle/123456789/116539
      Collections
      • UT - Forest Management [3207]

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository
        

       

      Browse

      All of IPB RepositoryCollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

      My Account

      Login

      Application

      google store

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository