使用深度学习来检测修剪区域和植物器官细分在休眠的刺修剪的葡萄树上
P Guadagna1, M Fernandes2, F Chen2
1Department of Sustainable Crop Production (DI.PRO.VE.S.), Università Cattolica del Sacro Cuore, Via Emilia Parmense 84, 29122 Piacenza, Italy.
Precision agriculture
|June 26, 2023
概括
深度学习模型进行了微调,以检测修剪区域和细分葡萄藤器官用于机器人冬季修剪. 人工智能驱动的树冠管理提高了这些自动化葡萄园解决方案的性能.
科学领域:
- 农业机器人农业机器人
- 计算机视觉 计算机视觉
- 葡萄种植 葡萄种植 葡萄种植
背景情况:
- 葡萄园管理成本仍然存在,原因是劳动密集型的选择性操作,如冬季修剪.
- 机器人解决方案正在农业中出现,但它们对葡萄树的应用仍未得到充分探索.
研究的目的:
- 微调和测试深度神经网络以检测修剪区域 (PRs) 和细分葡萄藤器官 (GO).
- 开发用于葡萄树冬季修剪的自主机器人系统的部件.
主要方法:
- 微调更快的R-CNN具有1215个RGB图像用于PR检测,在232张图像上进行测试.
- 训练面具R-CNN具有119个RGB图像用于GO细分 (带,臂,杆,杆,节点),在60张图像上进行了测试.
主要成果:
- 更快的R-CNN实现了高PR检测率,特别是在可见的杆 (0.97).
- 面具R-CNN展示了有效的GO细分,节点是最好的细分 (0.88).
- 发芽稀疏 (ST) 葡萄树的分段召回率高于对照组 (0.85),高于对照组 (0.80),表明了树冠管理的影响.
结论:
- 深度学习模型显示出对自动化葡萄藤冬季修剪任务的承诺.
- 能见度显著影响了修剪区域检测的准确性.
- 树冠管理策略可以提高基于AI的机器人解决方案在葡萄栽培中的性能.
相关概念视频
01:18Plant Tissues
6.5K
Plants are multicellular eukaryotes with tissue systems made of various cell types that carry out specific functions. Different tissues work together to perform a unique function and form an organ. Organs working together form organ systems. Vascular plants have two distinct organ systems: a shoot system and a root system. The shoot system consists of two portions: the vegetative (non-reproductive) parts of the plant, such as the leaves and the stems, and the reproductive parts of the plant,...
6.5K
02:27Basic Plant Anatomy: Roots, Stems, and Leaves
59.6K
The primary organs of vascular plants are roots, stems, and leaves, but these structures can be highly variable, adapted for the specific needs and environment of different plant species.
59.6K
03:24Seedless Vascular Plants
60.7K
Seedless Vascular Plants Were the First Tall Plants on Earth
60.7K