Field dataset and cabbage growth stage recognition for nutrient spraying robots
Published: August 11, 2026
Abstract
Accurate identification of cabbage growth stages is essential for precise fertilization, and its effectiveness depends directly on training data. This study constructs a field dataset and evaluates models for recognizing cabbage growth stages to support precision nutrient spraying robots. A total of 7,225 field images collected at Me Ha were combined with 1,786 UAV images, forming a dataset of 9,011 images with 16,616 annotations across five growth stages. Among five object detection models, YOLOv8n was selected due to its balance between accuracy (mAP 85.6%) and speed, achieving approximately 21 FPS on Jetson Nano after TensorRT FP16 optimization. Experimental results show over 90% recognition accuracy and spraying position deviation below 1 cm. The study contributes to dataset development and provides a suitable recognition solution for agricultural robotic systems, enabling the application of artificial intelligence in precision farming.
Andaloro J. T., Rose K. B., Shelton A. M., et al. (1983). Cabbage Growth Stages. New York State Agricultural Experiment Station, 101, 1-4.
2.
Howard A., Sandler M., Chu G., et al. (2019). Searching for MobileNetV3. arXiv:1905.02244.
3.
Jiang P., Qi A., Zhong J., et al. (2024). Field cabbage detection and positioning system based on improved YOLOv8n. Plant Methods, 20, 96.
4.
Lin T.-Y., Maire M., Belongie S., et al. (2014). Microsoft COCO: Common Objects in Context. arXiv:1405.0312.
5.
Liu W., Anguelov D., Erhan D., et al. (2015). SSD: Single Shot MultiBox Detector. arXiv:1512.02325.
6.
Lu Y., & Young S. (2020). A survey of public datasets for computer vision tasks in precision agriculture. Computers and Electronics in Agriculture, 178, 105760.
7.
Lüling N., Reiser D., Straub J., et al. (2023). Fruit Volume and Leaf-Area Determination of Cabbage by a Neural-Network-Based Instance Segmentation for Different Growth Stages. Sensors, 23, 129.
8.
Redmon J., Divvala S., Girshick R., et al. (2015). You Only Look Once: Unified, Real-Time Object Detection. arXiv:1506.02640.
9.
Ren S., He K., Girshick R., et al. (2015). Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. arXiv:1506.01497.
Vatin N. I., Joshi S. K., Acharya P., et al. (2024). Precision Agriculture and Sustainable Yields: Insights from IoT-Driven Farming. BIO Web of Conferences, 86, 01091.
14.
Yokoyama Y., Matsui T., & Tanaka S. T. (2024). An annotated image dataset of cabbages for instance segmentation. Mendeley Data, V2. https://doi.org/10.17632/5cp2dyjczk.2.
