Rancang Bangun Sistem Alat Handheld Untuk Klasifikasi Biji Jagung Menggunakan Image Processing Dengan Metode YOLOv8-PSO

Zidan Hilmi, Muhammad Rafi (2026) Rancang Bangun Sistem Alat Handheld Untuk Klasifikasi Biji Jagung Menggunakan Image Processing Dengan Metode YOLOv8-PSO. Diploma thesis, Institut Teknologi Sepuluh Nopember.

Text 2042221125-Undergraduate_Thesis.pdf Download (5MB)

Abstract

Klasifikasi biji jagung secara manual cenderung memakan waktu dan tidak selalu konsisten. Penelitian ini merancang Alat Handheld berbasis image processing untuk mengklasifikasikan biji jagung menjadi dua kelas, yaitu Good dan Bad, menggunakan YOLOv8m yang dioptimasi Particle Swarm Optimization (PSO) pada hyperparameter. Sistem dilengkapi kamera, pencahayaan LED, tray grid dengan ROI, dan HMI untuk menampilkan hasil secara real-time. Dataset yang digunakan berjumlah 1.702 citra. Pengujian dilakukan pada beberapa skenario tray dengan variasi warna tray dan komposisi sampel, menggunakan metrik precision, recall, F1-Score, mAP, serta Confution Matrix. Hasil menunjukkan YOLOv8m+PSO lebih baik dibandingkan baseline YOLOv8m, dengan F1-Score 0.96, akurasi 0.93 dan waktu proses 7 detik, sedangkan baseline memperoleh F1-Score 0.92, akurasi 0.86, dan waktu proses 10 detik. Peningkatan juga terlihat dari naiknya TP dan turunnya FN.

==================================================================================================================================

Manual classification of corn kernels tends to be time-consuming and inconsistent. This study designed a handheld image processing device to classify corn kernels into two classes, namely Good and Bad, using YOLOv8m optimised by Particle Swarm Optimization (PSO) on hyperparameters. The system is equipped with a camera, LED lighting, a grid tray with ROI, and an HMI to display results in real-time. The dataset used consists of 1,702 images. Testing was conducted on several tray scenarios with variations in tray colour and sample composition, using precision, recall, F1-Score, mAP, and Confusion Matrix metrics. The results show that YOLOv8m+PSO is better than the baseline YOLOv8m, with an F1-Score of 0.96, accuracy of 0.93, and processing time of 7 seconds, while the baseline obtained an F1-Score of 0.92, accuracy of 0.86, and processing time of 10 seconds. Improvements were also seen in the increase in TP and decrease in FN.

Item Type:	Thesis (Diploma)
Uncontrolled Keywords:	biji jagung, handheld, YOLOv8m, Particle Swarm Optimization, klasifikasi, real-time. corn kernels, handheld, YOLOv8m, Particle Swarm Optimisation, classification, real-time.
Subjects:	S Agriculture > SB Plant culture > SB409.58 Plant propagation. Including in vitro propagation T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.62 Simulation T Technology > T Technology (General) > T59.7 Human-machine systems. T Technology > TS Manufactures > TS171 Product design
Divisions:	Faculty of Vocational > Instrumentation Engineering
Depositing User:	Muhammad Rafi Zidan Hilmi
Date Deposited:	31 Jul 2026 07:30
Last Modified:	31 Jul 2026 07:30
URI:	http://repository.its.ac.id/id/eprint/140897

Actions (login required)

View Item