Sistem Deteksi Otomatis Defisiensi Nutrisi Bibit Kelapa Sawit Berbasis CNN MobileNetV2

Al Dhavi, Reinra (2026) Sistem Deteksi Otomatis Defisiensi Nutrisi Bibit Kelapa Sawit Berbasis CNN MobileNetV2. Bachelor thesis, Institut Teknologi Kalimantan.

[img] Text
04221031_cover.pdf
Restricted to Repository staff only until 4 October 2028.
Download (137kB) | Request a copy
[img] Text
04221031_statement_of_authenticity.pdf
Restricted to Repository staff only until 4 October 2028.
Download (147kB) | Request a copy
[img] Text
04221031_publishing_agreement.pdf
Restricted to Repository staff only until 4 October 2028.
Download (139kB) | Request a copy
[img] Text
04221031_approval_sheet.pdf
Restricted to Repository staff only until 4 October 2028.
Download (123kB) | Request a copy
[img] Text
04221031_preface.pdf
Restricted to Repository staff only until 4 October 2028.
Download (312kB) | Request a copy
[img] Text
04221031_abstract_id.pdf
Restricted to Repository staff only until 4 October 2028.
Download (385kB) | Request a copy
[img] Text
04221031_abstract_en.pdf
Restricted to Repository staff only until 4 October 2028.
Download (357kB) | Request a copy
[img] Text
04221031_table_of_content.pdf
Restricted to Repository staff only until 4 October 2028.
Download (335kB) | Request a copy
[img] Text
04221031_illustrations.pdf
Restricted to Repository staff only until 4 October 2028.
Download (266kB) | Request a copy
[img] Text
04221031_tables.pdf
Restricted to Repository staff only until 4 October 2028.
Download (313kB) | Request a copy
[img] Text
04221031_notations.pdf
Restricted to Repository staff only until 4 October 2028.
Download (313kB) | Request a copy
[img] Text
04221031_chapter_1.pdf
Restricted to Repository staff only until 4 October 2028.
Download (261kB) | Request a copy
[img] Text
04221031_chapter_2.pdf
Restricted to Repository staff only until 4 October 2028.
Download (430kB) | Request a copy
[img] Text
04221031_chapter_3.pdf
Restricted to Repository staff only until 4 October 2028.
Download (365kB) | Request a copy
[img] Text
04221031_chapter_4.pdf
Restricted to Repository staff only until 4 October 2028.
Download (1MB) | Request a copy
[img] Text
04221031_conclusions.pdf
Restricted to Repository staff only until 4 October 2028.
Download (196kB) | Request a copy
[img] Text
04221031_bibliography.pdf
Restricted to Repository staff only until 4 October 2028.
Download (181kB) | Request a copy
[img] Text
04221031_enclosure.pdf
Restricted to Repository staff only until 4 October 2028.
Download (696kB) | Request a copy
[img] Text
04221031_paper.pdf
Restricted to Repository staff only until 4 October 2028.
Download (579kB) | Request a copy
[img] Text
04221031_Form. TA-020.pdf
Restricted to Repository staff only until 4 October 2028.
Download (381kB) | Request a copy
[img] Text
04221031_presentation.pdf
Restricted to Repository staff only until 4 October 2028.
Download (1MB) | Request a copy
[img] Text
04221031_presentation.pdf
Download (1MB)

Abstract

Kelapa sawit (Elaeis guineensis Jacq.) merupakan komoditas perkebunan strategis yang berperan penting dalam perekonomian Indonesia, sehingga kualitas bibit pada tahap pembibitan (nursery) menjadi faktor krusial dalam menentukan produktivitas tanaman di masa mendatang. Salah satu permasalahan utama pada fase pembibitan adalah terjadinya defisiensi nutrisi, khususnya nitrogen (N), fosfor (P), kalium (K), dan magnesium (Mg), yang umumnya ditandai dengan perubahan warna daun. Identifikasi kondisi defisiensi nutrisi di lapangan hingga saat ini masih banyak dilakukan secara manual melalui pengamatan visual oleh tenaga kerja, sehingga bersifat subjektif, kurang konsisten, serta dipengaruhi oleh fluktuasi pencahayaan. Oleh karena itu, penelitian ini bertujuan untuk merancang dan membangun sistem deteksi otomatis defisiensi nutrisi pada bibit kelapa sawit berbasis analisis citra daun menggunakan metode Convolutional Neural Network (CNN) dengan arsitektur MobileNetV2. Dataset yang digunakan terdiri dari 300 citra daun bibit kelapa sawit yang dibagi ke dalam lima kategori nutrisi (Normal, defisiensi N, P, K, dan Mg). Data citra diproses melalui tahapan pre-processing yang meliputi pemotongan Region of Interest (ROI), normalisasi warna, resize menjadi 224×224 piksel, dan augmentasi citra, kemudian dibagi menjadi 80% data latih dan 20% data uji. Model dilatih menggunakan pendekatan transfer learning dengan bobot ImageNet dan dievaluasi melalui penyetelan hyperparameter. Hasil pengujian menunjukkan bahwa konfigurasi paling optimal dicapai pada Epoch 100, Learning Rate 0,000001, dan Batch Size 16, yang menghasilkan akurasi keseluruhan sebesar 95,38% dan rata-rata F1-Score 0,95. Bobot model terbaik kemudian diimplementasikan ke dalam prototipe aplikasi mobile bernama "SMART DETECT" yang mampu mendeteksi kondisi daun secara real-time sekaligus memberikan rekomendasi penanganan. Sistem ini terbukti efektif dan andal untuk dimanfaatkan sebagai sistem pendukung keputusan (decision support) dalam penerapan teknologi smart agriculture di sektor pembibitan kelapa sawit.

Item Type: Thesis (Bachelor)
Subjects: T Technology > T Technology (General)
T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions: Jurusan Teknologi Industri dan Proses > Teknik Elektro
Depositing User: Reinra Al Dhavi
Date Deposited: 13 Jul 2026 06:21
Last Modified: 13 Jul 2026 06:21
URI: http://repository.itk.ac.id/id/eprint/26783

Actions (login required)

View Item View Item