Real-time and resource-efficient banana bunch detection and localization with YOLO-BRFB on edge devices
Reliable detection and spatial localization of banana bunches are essential prerequisites for the development of autonomous harvesting technologies. Current methods face challenges in achieving high detection accuracy and efficient deployment due to their structural complexity and significant comput...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-08-01
|
| Series: | Frontiers in Plant Science |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fpls.2025.1650012/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849231598356004864 |
|---|---|
| author | Shuo Wang Shuo Wang Shuo Wang Lijiao Wei Danran Zhang Danran Zhang Ling Chen Ling Chen Weihua Huang Dongjie Du Kangmin Lin Zhenhui Zheng Zhenhui Zheng Zhenhui Zheng Jieli Duan |
| author_facet | Shuo Wang Shuo Wang Shuo Wang Lijiao Wei Danran Zhang Danran Zhang Ling Chen Ling Chen Weihua Huang Dongjie Du Kangmin Lin Zhenhui Zheng Zhenhui Zheng Zhenhui Zheng Jieli Duan |
| author_sort | Shuo Wang |
| collection | DOAJ |
| description | Reliable detection and spatial localization of banana bunches are essential prerequisites for the development of autonomous harvesting technologies. Current methods face challenges in achieving high detection accuracy and efficient deployment due to their structural complexity and significant computational demands. This study proposes YOLO-BRFB, a lightweight and precise system designed for detection and 3D localization of bananas in orchard environments. First, the YOLOv8 framework is improved by integrating the BasicRFB module, enhancing feature extraction for small targets and cluttered backgrounds while reducing model complexity. Then, a binocular vision system is used for localization, estimating 3D spatial coordinates with high accuracy and ensuring robust performance under diverse lighting and occlusion conditions. Finally, the system is optimized for edge-device deployment, achieving real-time processing with minimal computational resources. Experimental results demonstrate that YOLO-BRFB achieves a precision of 0.957, recall of 0.922, mAP of 0.961, and F1-score of 0.939, surpassing YOLOv8 in both recall and mAP. The average positioning error of the system along the X-axis is 12.33 mm, the average positioning error along the Y-axis is 11.11 mm, and the average positioning error along the Z-axis is 16.33 mm. The system has an inference time of 8.6 milliseconds on an Nvidia Orin NX with a GPU memory requirement of 1.7 GB. This study is among the first to focus on a lightweight approach optimized for deployment on edge computing devices. These results highlight the practical applicability of YOLO-BRFB in real-world agricultural scenarios, providing a cost-effective solution for precision harvesting. |
| format | Article |
| id | doaj-art-da0b8bfb5eae4c6191146dd3d5d96c9f |
| institution | Kabale University |
| issn | 1664-462X |
| language | English |
| publishDate | 2025-08-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Plant Science |
| spelling | doaj-art-da0b8bfb5eae4c6191146dd3d5d96c9f2025-08-21T05:27:19ZengFrontiers Media S.A.Frontiers in Plant Science1664-462X2025-08-011610.3389/fpls.2025.16500121650012Real-time and resource-efficient banana bunch detection and localization with YOLO-BRFB on edge devicesShuo Wang0Shuo Wang1Shuo Wang2Lijiao Wei3Danran Zhang4Danran Zhang5Ling Chen6Ling Chen7Weihua Huang8Dongjie Du9Kangmin Lin10Zhenhui Zheng11Zhenhui Zheng12Zhenhui Zheng13Jieli Duan14Agricultural Machinery Research Institute, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaKey Laboratory of Agricultural Equipment for Tropical Crops, Ministry of Agriculture and Rural Affairs, Guangdong, ChinaInstitute of Agricultural Machinery, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaAgricultural Machinery Research Institute, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaAgricultural Machinery Research Institute, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaCollege of Engineering, South China Agricultural University, Guangdong, ChinaAgricultural Machinery Research Institute, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaCollege of Mathematics and Informatics, South China Agricultural University, Guangdong, ChinaAgricultural Machinery Research Institute, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaAgricultural Machinery Research Institute, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaCollege of Engineering, South China Agricultural University, Guangdong, ChinaAgricultural Machinery Research Institute, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaKey Laboratory of Agricultural Equipment for Tropical Crops, Ministry of Agriculture and Rural Affairs, Guangdong, ChinaInstitute of Agricultural Machinery, Chinese Academy of Tropical Agricultural Sciences, Guangdong, ChinaCollege of Engineering, South China Agricultural University, Guangdong, ChinaReliable detection and spatial localization of banana bunches are essential prerequisites for the development of autonomous harvesting technologies. Current methods face challenges in achieving high detection accuracy and efficient deployment due to their structural complexity and significant computational demands. This study proposes YOLO-BRFB, a lightweight and precise system designed for detection and 3D localization of bananas in orchard environments. First, the YOLOv8 framework is improved by integrating the BasicRFB module, enhancing feature extraction for small targets and cluttered backgrounds while reducing model complexity. Then, a binocular vision system is used for localization, estimating 3D spatial coordinates with high accuracy and ensuring robust performance under diverse lighting and occlusion conditions. Finally, the system is optimized for edge-device deployment, achieving real-time processing with minimal computational resources. Experimental results demonstrate that YOLO-BRFB achieves a precision of 0.957, recall of 0.922, mAP of 0.961, and F1-score of 0.939, surpassing YOLOv8 in both recall and mAP. The average positioning error of the system along the X-axis is 12.33 mm, the average positioning error along the Y-axis is 11.11 mm, and the average positioning error along the Z-axis is 16.33 mm. The system has an inference time of 8.6 milliseconds on an Nvidia Orin NX with a GPU memory requirement of 1.7 GB. This study is among the first to focus on a lightweight approach optimized for deployment on edge computing devices. These results highlight the practical applicability of YOLO-BRFB in real-world agricultural scenarios, providing a cost-effective solution for precision harvesting.https://www.frontiersin.org/articles/10.3389/fpls.2025.1650012/fullmachine visiondetection and localizationbanana buncheslightweight modeledge computing |
| spellingShingle | Shuo Wang Shuo Wang Shuo Wang Lijiao Wei Danran Zhang Danran Zhang Ling Chen Ling Chen Weihua Huang Dongjie Du Kangmin Lin Zhenhui Zheng Zhenhui Zheng Zhenhui Zheng Jieli Duan Real-time and resource-efficient banana bunch detection and localization with YOLO-BRFB on edge devices Frontiers in Plant Science machine vision detection and localization banana bunches lightweight model edge computing |
| title | Real-time and resource-efficient banana bunch detection and localization with YOLO-BRFB on edge devices |
| title_full | Real-time and resource-efficient banana bunch detection and localization with YOLO-BRFB on edge devices |
| title_fullStr | Real-time and resource-efficient banana bunch detection and localization with YOLO-BRFB on edge devices |
| title_full_unstemmed | Real-time and resource-efficient banana bunch detection and localization with YOLO-BRFB on edge devices |
| title_short | Real-time and resource-efficient banana bunch detection and localization with YOLO-BRFB on edge devices |
| title_sort | real time and resource efficient banana bunch detection and localization with yolo brfb on edge devices |
| topic | machine vision detection and localization banana bunches lightweight model edge computing |
| url | https://www.frontiersin.org/articles/10.3389/fpls.2025.1650012/full |
| work_keys_str_mv | AT shuowang realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT shuowang realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT shuowang realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT lijiaowei realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT danranzhang realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT danranzhang realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT lingchen realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT lingchen realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT weihuahuang realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT dongjiedu realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT kangminlin realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT zhenhuizheng realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT zhenhuizheng realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT zhenhuizheng realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices AT jieliduan realtimeandresourceefficientbananabunchdetectionandlocalizationwithyolobrfbonedgedevices |