CD Skripsi
Pengaruh Kecepatan Konveyor Dan Frame Rate Kamera Terhadap Performa Deteksi Tingkat Kematangan Tandan Buah Segar Kelapa Sawit
A machine vision system has been developed to assist in the automatic and real-time sorting and grading of oil palm Fresh Fruit Bunches (FFB), especially for detecting their ripeness level. The system consists of a computer vision unit, a conveyor, a control circuit, and a separation arm. The conveyor speed and the camera's frame rate per second (FPS) influence the accuracy of image detection. This study aims to analyze the effect of varying conveyor speeds and camera FPS on the performance of a computer vision-based detection system for oil palm FFB ripeness levels using the YOLOv8 algorithm. Two conveyor speeds were applied, namely 9.746 cm/s and 18.45 cm/s, along with three FPS variations: 10 fps, 20 fps, and 39.55 fps. Each combination of speed and FPS was tested to evaluate the model’s detection performance. The model’s performance was analyzed using a confusion matrix with evaluation parameters including accuracy, precision, recall, and F1-score. The results showed that the combination of conveyor speeds (9.746 cm/s and 18.45 cm/s) at 39.55 FPS provided the best detection performance, achieving 100% accuracy, precision, recall, and F1-score for both ripe and unripe FFB. This system shows potential in supporting the automation of FFB sorting processes in the palm oil industry in a faster, more accurate, and efficient manner.
Keywords : machine vision, oil palm fresh fruit bunch, conveyor speed, frame rate, YOLOv8
Tidak tersedia versi lain