RANCANG BANGUN SISTEM PERHITUNGAN DAN KLASIFIKASI KELAPA SAWIT BERBASIS YOLOV8 DENGAN INTEGRASI TEKNOLOGI IOT

Julio Caesar Ray Bakar Gani, - and Anugrah Adiwilaga, - and Devi Aprianti Rimadhani Agustini, - (2025) RANCANG BANGUN SISTEM PERHITUNGAN DAN KLASIFIKASI KELAPA SAWIT BERBASIS YOLOV8 DENGAN INTEGRASI TEKNOLOGI IOT. S1 thesis, Universitas Pendidikan Indonesia.

Abstract

Industri kelapa sawit merupakan salah satu sektor yang berkontribusi besar terhadap perekonomian nasional di Indonesia. Penentuan tingkat kematangan buah kelapa sawit merupakan aspek krusial dalam proses panen guna memastikan hasil produksi yang optimal. Penelitian ini bertujuan merancang dan membangun sistem otomatisasi perhitungan kelapa sawit berbasis model computer vision YOLOv8 yang terintegrasi dengan teknologi IoT serta evaluasi dari kinerja sistem. Metode penelitian yang digunakan mengikuti pendekatan Design and Development (D&D) dengan tahapan: identifikasi kebutuhan sistem, perancangan sistem, implementasi, dan evaluasi. SBC yang digunakan pada penelitian ini yaitu Raspberry Pi 5 dan mikrokontroler berupa esp32. Dataset yang digunakan terdiri dari dua kelas, yaitu matang dan mentah. Dataset diperoleh dari sumber publik dan dilakukan pelabelan ulang sebelum dilatih menggunakan model YOLOv8. Sistem diimplementasikan untuk mendeteksi buah sawit secara real-time melalui kamera, kemudian hasil deteksi ditampilkan pada website dan panel LED P10 melalui ESP32. Pengujian dilakukan untuk mengevaluasi akurasi model dan kinerja sistem. Hasil pelatihan menunjukkan bahwa model mencapai akurasi mAP50 sebesar 0.975 dan mAP50-95 sebesar 0.781. Dengan perangkat Raspberry Pi 5, diperoleh kecepatan pemrosesan sekitar 2 FPS. Sistem berhasil dirancang dan dibangun serta menunjukkan kinerja deteksi dengan akurasi rata-rata sebesar 85% pada pencahayaan terang, 80% pada pencahayaan normal, dan 78% pada pencahayaan redup dengan berbagai jarak pengujian. -------- The palm oil industry is one of the sectors that significantly contributes to Indonesia’s national economy. Determining the ripeness level of oil palm fruit is a crucial aspect of the harvesting process to ensure optimal production yields. This study aims to design and develop an automated palm oil fruit counting system based on the YOLOv8 computer vision model integrated with IoT technology, as well as to evaluate the system’s performance. The research method follows the Design and Development (D&D) approach, consisting of the following stages: system requirements identification, system design, implementation, and evaluation. The SBC used in this study is the Raspberry Pi 5, while the microcontroller utilized is the ESP32. The dataset consists of two classes, namely ripe and unripe. The dataset was obtained from a public source and relabeled before being trained using the YOLOv8 model. The system was implemented to detect oil palm fruits in real-time using a camera, with detection results displayed on a website and an LED P10 panel through the ESP32. Testing was conducted to evaluate the model accuracy and system performance. The training results show that the model achieved an mAP50 accuracy of 0.975 and an mAP50-95 of 0.781. Using the Raspberry Pi 5, the processing speed reached approximately 2 FPS. The system was successfully designed and developed, demonstrating detection performance with an average accuracy of 85% under bright lighting, 80% under normal lighting, and 78% under dim lighting across various testing distances.

	Text S_TEKOM_2103066_Title.pdf Download (5MB)
	Text S_TEKOM_2103066_Chapter1.pdf Download (1MB)
	Text S_TEKOM_2103066_Chapter2.pdf Restricted to Staf Perpustakaan Download (4MB)
	Text S_TEKOM_2103066_Chapter3.pdf Download (8MB)
	Text S_TEKOM_2103066_Chapter4.pdf Restricted to Staf Perpustakaan Download (8MB)
	Text S_TEKOM_2103066_Chapter5.pdf Download (563kB)
	Text S_TEKOM_2103066_Appendix.pdf Restricted to Staf Perpustakaan Download (3MB)

Item Type:	Thesis (S1)
Additional Information:	https://scholar.google.com/citations?view_op=list_works&hl=en&authuser=4&user=ZxI2lvcAAAAJ ID SINTA Dosen Pembimbing Anugrah Adiwilaga: 6745914 Devi Aprianti Rimadhani Agustini: 6745751
Uncontrolled Keywords:	Kelapa sawit, YOLOv8, Computer Vision, ESP32, otomatisasi perhitungan, Palm oil, Computer Vision, automated counting
Subjects:	L Education > L Education (General) Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software T Technology > T Technology (General)
Divisions:	UPI Kampus cibiru > S1 Teknik Komputer
Depositing User:	Julio Caesar Ray Bakar Gani
Date Deposited:	18 Sep 2025 06:57
Last Modified:	18 Sep 2025 06:57
URI:	http://repository.upi.edu/id/eprint/137564

Actions (login required)

View Item