View Item 
      •   IPB Repository
      • Dissertations and Theses
      • Undergraduate Theses
      • UT - Faculty of Agriculture
      • UT - Soil Science and Land Resources
      • View Item
      •   IPB Repository
      • Dissertations and Theses
      • Undergraduate Theses
      • UT - Faculty of Agriculture
      • UT - Soil Science and Land Resources
      • View Item
      JavaScript is disabled for your browser. Some features of this site may not work without it.

      Pemetaan Gulma Berbasis Object-Based Image Analysis Dengan Pendekatan Machine Learning Menggunakan Citra Drone Multispektral

      Thumbnail
      View/Open
      Cover (452.7Kb)
      Fulltext (4.556Mb)
      Lampiran (1.038Mb)
      Date
      2025
      Author
      Ramadhani, Rasya Aulia
      Ardiansyah, Muhammad
      Munibah, Khursatul
      Metadata
      Show full item record
      Abstract
      Penurunan produksi padi nasional sebesar 1,55% pada tahun 2024 menegaskan urgensi peningkatan produktivitas secara berkelanjutan, salah satunya melalui pengendalian gulma yang efektif. Identifikasi gulma menggunakan drone dengan kamera multispektral menawarkan pendekatan presisi tinggi dalam membedakan tanaman padi dan gulma berdasarkan analisis pantulan spektral. Penelitian ini bertujuan mengidentifikasi pola reflektansi dari gulma dan non-gulma, mengevaluasi parameter segmentasi Object-Based Image Analysis (OBIA) dan klasifikasi berbasis objek menggunakan metode RF dan SVM dalam identifikasi gulma di lahan sawah. Segmentasi dilakukan menggunakan dua pendekatan, yaitu Original Multiresolution (OMN) dan Region Grow on Object (RGO), dengan beberapa kombinasi parameter scale, shape, dan compactness. Hasil analisis pola reflektan menunjukkan perbedaan karakteristik spektral terutama pada kanal inframerah dekat antara tanaman padi dan gulma. Evaluasi segmentasi OBIA menunjukkan parameter scale berpengaruh paling dominan terhadap jumlah dan ukuran objek yang dihasilkan. Pendekatan OMN sensitivitas tinggi terhadap variasi piksel sehingga menghasilkan segmentasi yang lebih detail, sedangkan RGO cenderung menghasilkan segmentasi yang lebih umum dengan objek berukuran besar dan jumlah objek lebih sedikit. Dalam klasifikasi gulma, metode SVM menunjukkan kinerja yang lebih stabil dan akurasi yang lebih tinggi dibandingkan metode RF. Secara keseluruhan, kedua metode klasifikasi tersebut memiliki kemampuan yang baik dalam pemetaan persebaran gulma secara cepat.
       
      The decline in national rice production by 1.55% by 2024 emphasizes the urgency of increasing productivity in a sustainable manner, one of which is through effective weed control. Weed identification using drones with multispectral cameras offers a high-precision approach in distinguishing rice plants and weeds based on spectral reflectance analysis. This research aims to identify reflectance patterns of weeds and non-weeds, evaluate Object-Based Image Analysis (OBIA) segmentation parameters and object-based classification using RF and SVM methods in weed identification in paddy fields. Segmentation using two approach, Original Multiresolution (OMN) and Region Grow on Object (RGO), with various combinations of scale, shape, and compactness parameters. The results of reflectance pattern show differences in spectral characteristics, especially in the NIR (Near-Infrared) band between rice plants and weeds. OBIA results show that the scale parameter has the most significant effect on the number and size of objects. The OMN approach is highly sensitive to pixel variations resulting more detailed segmentation, while RGO tends to produce more generalized segmentation with large objects and fewer in number. In weed classification, the SVM method showed more stable performance and higher accuracy than the RF method. Overall, both classification methods performed well in mapping weed distribution quickly.
       
      URI
      http://repository.ipb.ac.id/handle/123456789/169038
      Collections
      • UT - Soil Science and Land Resources [2822]

      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