Saputra, Naufal Ammar (2025) Pengembangan Aplikasi Android Berbasis Machine Learning Untuk Deteksi Penyakit pada Bawang Merah yang Terintegrasi dengan Sistem Internet of Things. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5027211052_Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5027211052_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy |
Abstract
Bawang merah merupakan salah satu komoditas holtikultura yang memiliki nilai ekonomi tinggi di berbagai negara, termasuk indonesia, namun rentan terhadap berbagai penyakit seperti bercak ungu (Alternaria porri) dan layu fusarium atau moler (Fusarium oxysporum). Penelitian ini bertujuan untuk mengembangkan aplikasi Android berbasis machine learning yang terintegrasi dengan sistem Internet of Things (IoT) guna mendeteksi penyakit pada tanaman bawang merah secara dini. Model klasifikasi citra dalam penelitian ini dibangun menggunakan arsitektur MobileNetV2, yang dikenal efisien untuk implementasi pada perangkat mobile. Aplikasi diuji dari segi fungsionalitas, kompatibilitas, dan penerimaan pengguna. Hasil uji menunjukkan tingkat keberhasilan 93,75% pada aspek user acceptance, serta kompatibilitas perangkat sebesar 100% di berbagai tipe Android. Pengujian model menghasilkan akurasi sebesar 88% pada data tes, dengan performa tinggi pada kelas Trotol dan Busuk Daun, namun presisi pada kelas Moler, Serangga, dan Invalid masih perlu ditingkatkan. Uji lapangan dan uji variasi latar menunjukkan penurunan akurasi akibat citra tanaman yang terekam kurang fokus karena kepadatan tanaman. Selain itu, masukan pengguna mengusulkan perbaikan pada tampilan input gambar, indikator proses, dan peningkatan akurasi klasifikasi citra tidak relevan. Secara keseluruhan, aplikasi dinilai layak digunakan dan memiliki potensi besar sebagai solusi praktis dan ramah pengguna dalam mendeteksi penyakit tanaman bawang merah, baik di tingkat petani maupun rumah tangga, dengan syarat dilakukan pengembangan berkelanjutan berdasarkan kondisi lapangan.
=======================================================================================================================================
Shallots are a high-value horticultural commodity in many countries, including Indonesia, but their productivity is often hindered by various diseases such as purple blotch (Alternaria porri) and fusarium wilt or moler (Fusarium oxysporum). This study aims to develop an Android-based application utilizing machine learning integrated with an Internet of Things (IoT) system for early detection of shallot plant diseases. The image classification model was built using the MobileNetV2 architecture, chosen for its efficiency on mobile devices. The application was evaluated based on functionality, device compatibility, and user acceptance. Results showed a user acceptance rate of 93.75%, and 100% compatibility across various Android devices. The classification model achieved an accuracy of 88% on test data, with strong performance in classes such as Trotol and Leaf Rot, although improvement is still needed for Moler, Insect, and Invalid classes, especially in terms of precision and sensitivity. Field tests revealed a noticeable drop in accuracy due to the dense planting conditions, which caused image capture to include multiple leaf segments and plant parts, reducing classification accuracy and plant height estimation. User feedback also highlighted the need for improvements in the image input interface, detection progress indicators, and irrelevant image filtering. Overall, the application is considered feasible and promising as an effective, user-friendly solution for shallot disease detection, benefiting both farmers and households, provided that continuous development is carried out based on real-world challenges.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Android, Machine Learning, Internet of Things (IoT), Deteksi Penyakit, Bawang Merah, MobileNetV2, Klasifikasi citra ================================================================== Android, Machine Learning, Internet of Things (IoT), Disease Detection, Shallots, MobileNetV2, Image classification |
| Subjects: | T Technology > T Technology (General) |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis |
| Depositing User: | Naufal Ammar Saputra |
| Date Deposited: | 04 Aug 2025 10:20 |
| Last Modified: | 04 Aug 2025 10:20 |
| URI: | http://repository.its.ac.id/id/eprint/124900 |
Actions (login required)
 |
View Item |