Please use this identifier to cite or link to this item: https://repositori.uma.ac.id/handle/123456789/18381
Full metadata record
DC FieldValueLanguage
dc.contributor.authorIfantiska, Dian-
dc.date.accessioned2022-11-16T07:46:02Z-
dc.date.available2022-11-16T07:46:02Z-
dc.date.issued2022-08-31-
dc.identifier.urihttps://repositori.uma.ac.id/handle/123456789/18381-
dc.description68 Halamanen_US
dc.description.abstractDalam aktivitas budidaya kelapa sawit para petani kerap kali mengalami bermacam-macam serbuan penyakit yang melanda Tanaman kelapa sawit. Minimnya wawasan petani sawit mengenai penyakit pada tanaman kelapa sawit jadi hambatan para petani akan menanggulangi sendiri penyakit yang melanda tanaman kelapa sawit yang dimiliki. Oleh karena itu dibuatlah penelitian untuk mengenali penyakit pada daun kelapa sawit khususnya pada daun yang terkena hama ulat api dan ulat kantung. Convulutional Neural Network (CNN) banyak digunakan pada penelitian terdahulu karena akurasinya yang tinggi dan memiliki beberapa pengembangan arsitektur, diantaranya terdapat arsitektur Googlenet dan Xception. Arsitektur ini dipilih karena memiliki tingkat akurasi yang tinggi pada ILSVRC (ImageNet Large-Scale Visual Recognition Challenge) dan menggunakan pendekatan transfer learning yang popular. Pada penelitian ini, Mengimplementasikan dan membandinkan Googlenet dan Xception untuk identifikasi penyakit pada daun tanaman kelapa sawit, dengan jumlah dataset sebanyak 1230 gambar , yang terdiri dari gambar daun sehat,daun yang terkena ulat api,dan daun yang terkena ulat kantung. Setelah data gambar penyakit daun dilatih, model data pelatihan akan disimpan untuk proses pengujian. Evaluasi pengujian disimpan sebagai confusion matriks. Hasilnya, Xception memiliki model akurasi sebesar 99,72 % sedangkan Googlenet memiliki model akurasi sebesar 93,22 %. In oil palm cultivation activities, farmers often experience various kinds of disease attacks that hit oil palm plants. The lack of knowledge of oil palm farmers about diseases in oil palm plants is an obstacle for farmers to overcome the diseases that affect their oil palm plants. Therefore, a study was made to identify diseases on oil palm leaves, especially on leaves affected by caterpillars and bagworms. Convultional Neural Network (CNN) is widely used in previous studies because of its high accuracy and has several architectural developments, including the Googlenet and Xception architectures. This architecture was chosen because it has a high level of accuracy in the ILSVRC (ImageNet Large-Scale Visual Recognition Challenge) and uses the popular transfer learning approach. In this study, implementing and comparing Googlenet and Xception to identify diseases on oil palm leaves, with a total dataset of 1230 images, consisting of images of healthy leaves, leaves affected by caterpillars, and leaves affected by bagworms. After the leaf disease image data is trained, the training data model will be saved for the testing process. The test evaluation is stored as a confusion matrix. As a result, Xception has an accuracy model of 99.72% while Googlenet has an accuracy model of 93.22%.en_US
dc.language.isootheren_US
dc.publisherUniversitas Medan Areaen_US
dc.relation.ispartofseriesNPM;178160083-
dc.subjectkelapa sawiten_US
dc.subjectgoogleneten_US
dc.subjectxceptionen_US
dc.subjectpalm oilen_US
dc.titleImplementasi Arsitektur Googlenet dan Xception untuk Identifikasi Penyakit pada Daun Tanaman Kelapa Sawiten_US
dc.title.alternativeImplementation of Googlenet Architecture and Xception for Identification of Diseases on Palm Oil Leavesen_US
dc.typeThesisen_US
Appears in Collections:SP - Informatic Engineering

Files in This Item:
File Description SizeFormat 
178160083 - Dian Ifantiska - Fulltext.pdfCover, Abstract, Chapter I, II, III, V, Bibliography2.37 MBAdobe PDFView/Open
178160083 - Dian Ifantiska - Chapter IV.pdf
  Restricted Access
Chapter IV1.16 MBAdobe PDFView/Open Request a copy


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.