Identification of the Maturity Level of Oil Palm Fruit Using a Combination of the You Only Look Once Version 8 Model and Convolutional Neural Network

DOI

10.2991/978-94-6463-825-7_13How to use a DOI?

Keywords

Palm oil fruit; Object detection model; Classification model; You Only Look Once version 8; Convolutional Neural Network

Abstract

The palm oil industry plays a vital role in Indonesia’s economy. One of the key challenges in maintaining high crude palm oil (CPO) quality is accurately determining the ripeness level of oil palm fruits. Innovative technologies have been established to identify the level of ripening of palm oil fruits. However, existing methods often struggle because of the complex visual characteristics of fruits, including noise, redundant information, and difficulty in extracting relevant features. This study proposes a combination of the You Only Look Once Version 8 (YOLOv8) model and a Convolutional Neural Network (CNN) to determine the ripeness level of oil palm fruits. The dataset used comprised 4,160 images containing 14,559 annotated objects across six categories: raw, underripe, ripe, overripe, abnormal, and empty bunch. Data were collected from an oil palm mill in South Kalimantan, Indonesia. Two training scenarios were conducted: the first evaluated the effect of different backbones (ResNet50, GhostNetP2, GhostNetP6, and DarkNet53), and the second examined the effect of hyperparameter tuning. The best-performing model, YOLOv8 with the DarkNet53 backbone, achieved a mean Average Precision (mAP) of 91.43% at an Intersection over Union (IoU) threshold of 0.50, along with an F1-score of 99.08%. Additionally, the model required only 2.99 hours for training over 163 epochs. These results highlight the effectiveness and efficiency of the proposed approach for automatic ripeness classification in palm-oil quality control.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Dwi Ratna Sulistyaningrum
AU  - Muchammad Aquila Azahari
AU  - Budi Setiyono
AU  - Dieky Adzkya
PY  - 2025
DA  - 2025/08/22
TI  - Identification of the Maturity Level of Oil Palm Fruit Using a Combination of the You Only Look Once Version 8 Model and Convolutional Neural Network
BT  - Proceedings of the International Conference of Tropical Agrifood Feed and Fuel 2024 (ICTAFF 2024)
PB  - Atlantis Press
SP  - 163
EP  - 177
SN  - 2468-5747
UR  - https://doi.org/10.2991/978-94-6463-825-7_13
DO  - 10.2991/978-94-6463-825-7_13
ID  - Sulistyaningrum2025
ER  -