YOLO-Ginseng: A Method for Detecting Ginseng Fruits
Author Information
Author(s): Xie Zhedong, Yang Zhuang, Li Chao, Zhang Zhen, Jiang Jiazhuo, Guo Hongyu
Primary Institution: College of Engineering and Technology, Jilin Agricultural University, Changchun, China
Hypothesis
Can the YOLO-Ginseng detection method improve the accuracy and speed of ginseng fruit detection in complex agricultural environments?
Conclusion
The YOLO-Ginseng method significantly enhances the detection performance of ginseng fruits, achieving high precision and speed.
Supporting Evidence
- YOLO-Ginseng achieved an average precision of 95.6%.
- The model inference time was recorded at 7.4 ms.
- Model weight size was reduced by 76.4% after compression.
- Detection speed reached 24.9 fps on the Jetson Orin Nano device.
Takeaway
This study created a new way to find ginseng fruits quickly and accurately, which can help in picking them more efficiently.
Methodology
The study used a deep learning model called YOLO-Ginseng, which includes a new feature extraction module and a model compression technique.
Limitations
The detection performance for long-distance ginseng fruits is still not ideal due to their small size and occlusion by other plants.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website