Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4
Fresh Fruit Bunch (FFB) is the main ingredient in palm oil production. Harvesting FFB from oil palm trees at its peak ripeness stage is crucial to maximise the oil extraction rate (OER) and quality. In current harvesting practices, misclassification of FFB ripeness can occur due to human error, resu...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Published: |
Institute of Electrical and Electronics Engineers
2022
|
| _version_ | 1811137634073837568 |
|---|---|
| author | Lai, Jin Wern Ramli, Hafiz Rashidi Ismail, Luthffi Idzhar Wan Hasan, Wan Zuha |
| author_facet | Lai, Jin Wern Ramli, Hafiz Rashidi Ismail, Luthffi Idzhar Wan Hasan, Wan Zuha |
| author_sort | Lai, Jin Wern |
| collection | UPM |
| description | Fresh Fruit Bunch (FFB) is the main ingredient in palm oil production. Harvesting FFB from oil palm trees at its peak ripeness stage is crucial to maximise the oil extraction rate (OER) and quality. In current harvesting practices, misclassification of FFB ripeness can occur due to human error, resulting in OER loss. Therefore, a vision-based ripe FFB detection system is proposed as the first step in a robotic FFB harvesting system. In this work, live camera input is fed into a Convolutional Neural Network (CNN) model known as YOLOv4 to detect the presence of ripe FFBs on the oil palm trees in real-time. Once a ripe FFB is detected on the tree, a signal is transmitted via ROS to the robotic harvesting mechanism. To train the YOLOv4 model, a large number of ripe FFB images were collected using an Intel Realsense Camera D435 with a resolution of 1920× 1080. During data acquisition, a subject matter expert assisted in classifying the FFBs in terms of ripe or unripe. During the testing phase, the result of the mean Average Precision (mAP) and recall are 87.9 % and 82 % as the detection fulfilled the Intersect over Union (IoU) with more than 0.5 after 2000 iterations and the system operated at the real-time speed of roughly 21 Frame Per Second (FPS). |
| first_indexed | 2024-09-25T03:37:25Z |
| format | Article |
| id | upm.eprints-102993 |
| institution | Universiti Putra Malaysia |
| last_indexed | 2024-09-25T03:37:25Z |
| publishDate | 2022 |
| publisher | Institute of Electrical and Electronics Engineers |
| record_format | dspace |
| spelling | upm.eprints-1029932024-06-30T06:57:06Z http://psasir.upm.edu.my/id/eprint/102993/ Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4 Lai, Jin Wern Ramli, Hafiz Rashidi Ismail, Luthffi Idzhar Wan Hasan, Wan Zuha Fresh Fruit Bunch (FFB) is the main ingredient in palm oil production. Harvesting FFB from oil palm trees at its peak ripeness stage is crucial to maximise the oil extraction rate (OER) and quality. In current harvesting practices, misclassification of FFB ripeness can occur due to human error, resulting in OER loss. Therefore, a vision-based ripe FFB detection system is proposed as the first step in a robotic FFB harvesting system. In this work, live camera input is fed into a Convolutional Neural Network (CNN) model known as YOLOv4 to detect the presence of ripe FFBs on the oil palm trees in real-time. Once a ripe FFB is detected on the tree, a signal is transmitted via ROS to the robotic harvesting mechanism. To train the YOLOv4 model, a large number of ripe FFB images were collected using an Intel Realsense Camera D435 with a resolution of 1920× 1080. During data acquisition, a subject matter expert assisted in classifying the FFBs in terms of ripe or unripe. During the testing phase, the result of the mean Average Precision (mAP) and recall are 87.9 % and 82 % as the detection fulfilled the Intersect over Union (IoU) with more than 0.5 after 2000 iterations and the system operated at the real-time speed of roughly 21 Frame Per Second (FPS). Institute of Electrical and Electronics Engineers 2022-09 Article PeerReviewed Lai, Jin Wern and Ramli, Hafiz Rashidi and Ismail, Luthffi Idzhar and Wan Hasan, Wan Zuha (2022) Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4. IEEE Access, 10. pp. 95763-95770. ISSN 2169-3536 https://ieeexplore.ieee.org/document/9878339/ 10.1109/access.2022.3204762 |
| spellingShingle | Lai, Jin Wern Ramli, Hafiz Rashidi Ismail, Luthffi Idzhar Wan Hasan, Wan Zuha Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4 |
| title | Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4 |
| title_full | Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4 |
| title_fullStr | Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4 |
| title_full_unstemmed | Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4 |
| title_short | Real-time detection of ripe oil palm fresh fruit bunch based on YOLOv4 |
| title_sort | real time detection of ripe oil palm fresh fruit bunch based on yolov4 |
| work_keys_str_mv | AT laijinwern realtimedetectionofripeoilpalmfreshfruitbunchbasedonyolov4 AT ramlihafizrashidi realtimedetectionofripeoilpalmfreshfruitbunchbasedonyolov4 AT ismailluthffiidzhar realtimedetectionofripeoilpalmfreshfruitbunchbasedonyolov4 AT wanhasanwanzuha realtimedetectionofripeoilpalmfreshfruitbunchbasedonyolov4 |