KaraAgroAI/Drone-based-Agricultural-Dataset-for-Crop-Yield-Estimation
收藏资源简介:
该数据集包含由无人机拍摄的腰果、可可和咖啡图像,每张图像都配有YOLO格式的标注文件,用于目标检测。数据集由加纳(KaraAgro AI)和乌干达(Makerere AI Lab, Uganda Marconi Lab, National Coffee Research Institute, National Crops Resources Research Institute)的团队收集,旨在为作物产量估计模型的开发和评估提供支持。数据集包含不同作物、生长阶段和环境条件的图像,以帮助研究人员开发适应现实农业场景的模型。加纳数据集包含8,784张图像,乌干达数据集包含6,086张图像。数据集的结构和标签信息也在README中详细列出。
This dataset contains drone-captured images of cashews, cocoa, and coffee, with each image paired with annotation files in YOLO format for object detection. The dataset was collected by teams from Ghana (KaraAgro AI) and Uganda (Makerere AI Lab, Uganda Marconi Lab, National Coffee Research Institute, National Crops Resources Research Institute), aiming to support the development and evaluation of crop yield estimation models. It includes images covering different crops, growth stages and environmental conditions to help researchers develop models adaptable to real-world agricultural scenarios. The Ghanaian subset contains 8,784 images, while the Ugandan subset contains 6,086 images. The dataset structure and label information are also detailed in the README file.
数据集概述
数据集名称
- Drone-based Agricultural Dataset for Crop Yield Estimation
数据集来源
- 由加纳的KaraAgro AI和乌干达的Makerere AI Lab, Uganda Marconi Lab, National Coffee Research Institute, National Crops Resources Research Institute团队收集。
数据集内容
- 包含腰果、可可和咖啡的无人机图像,每张图像配有YOLO格式的标注文件,包含类别标签和边界框坐标。
数据集规模
- 加纳部分:共有8,784张图像,分辨率为16000x13000像素。
- 腰果:4,715张图像
- 可可:4,069张图像
- 乌干达部分:共有6,086张图像,分辨率为4000x3000像素。
- 腰果:3,086张图像
- 咖啡:3,000张图像
数据集标签
- 腰果:[cashew_tree, flower, immature, mature, ripe, spoilt]
- 可可:[cocoa-pod-mature-unripe, cocoa-tree, cocoa-pod-immature, cocoa-pod-riped, cocoa-pod-spoilt]
- 咖啡:[unripe, ripening, ripe, spoilt, coffee]
数据集结构
Data/ └── Ghana/ ├── cashew.zip ├── cocoa.zip └── Uganda/ ├── cashew.zip ├── coffee.zip
数据集用途
- 主要用于作物产量估计,也可用于进一步的研究,如作物异常检测。
数据集许可证
- cc-by-4.0
数据集指标
- mape
数据集标签详细信息
- 腰果:
- 实例数量:{cashew_tree:1107, flower:16757, immature:11766, mature: 4244, ripe: 11721, spoilt: 518}
- 可可:
- 实例数量:{cocoa-pod-mature-unripe: 10786, cocoa-tree: 2831, cocoa-pod-immature: 2401, cocoa-pod-riped: 4193, cocoa-pod-spoilt: 2018}
数据集创建信息
- 由KaraAgro AI基金会的数据科学家团队创建,得到农业科学家和官员的支持,并由Lacuna Fund资助。




