Prakhar047/CADI-AI

---
license: cc-by-sa-4.0
task_categories:
- object-detection
language:
- en
tags:
- object detection
- vision
size_categories:
- 1K<n<10K

extra_gated_heading: "Acknowledge license to accept the repository"
extra_gated_button_content: "Acknowledge license"
extra_gated_fields:
I agree to attribute the creator of this repository: checkbox
---

---
## Cashew Disease Identication with Artificial Intelligence (CADI-AI) Dataset

This repository contains a comprehensive dataset of cashew images captured by drones, accompanied by meticulously annotated labels.
Each high-resolution image in the dataset has a resolution of 1600x1300 pixels, providing fine details for analysis and model training.
To facilitate efficient object detection, each image is paired with a corresponding text file in YOLO format.
The YOLO format file contains annotations, including class labels and bounding box coordinates.

### Dataset Labels

```
['abiotic', 'insect', 'disease']
```

### Number of Images

```json
{'train': 3788, 'valid': 710, 'test': 238}
```

### Number of Instances Annotated

```json
{'insect':1618, 'abiotic':13960, 'disease':7032}
```
### Folder structure after unzipping repective folders

```markdown
Data/
└── train/
├── images
├── labels
└── val/
├── images
├── labels
└── test/
├── images
├── labels
```

### Dataset Information
The dataset was created by a team of data scientists from the KaraAgro AI Foundation,
with support from agricultural scientists and officers.
The creation of this dataset was made possible through funding of the
Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) through their projects
[Market-Oriented Value Chains for Jobs & Growth in the ECOWAS Region (MOVE)](https://www.giz.de/en/worldwide/108524.html) and
[FAIR Forward - Artificial Intelligence for All](https://www.bmz-digital.global/en/overview-of-initiatives/fair-forward/), which GIZ implements on
behalf the German Federal Ministry for Economic Cooperation and Development (BMZ).

For detailed information regarding the dataset, we invite you to explore the accompanying datasheet available [here](https://drive.google.com/file/d/1viv-PtZC_j9S_K1mPl4R1lFRKxoFlR_M/view?usp=sharing).
This comprehensive resource offers a deeper understanding of the dataset's composition, variables, data collection methodologies, and other relevant details.

许可证：CC BY-SA 4.0
任务类别：目标检测
语言：英语
标签：目标检测、视觉
样本规模：1000 < 样本量 < 10000

额外访问提示标题："确认许可协议以获取本仓库访问权限"
访问按钮文本："确认许可"
额外访问字段："我同意标注本仓库的创建者：复选框"

## 基于人工智能的腰果病害识别（CADI-AI）数据集

本仓库收录了一套由无人机采集的腰果图像综合数据集，附带经过精细标注的标签。数据集内每张高分辨率图像的分辨率均为1600×1300像素，可为分析与模型训练提供充足的细节信息。为适配高效的目标检测任务，每张图像均配套一个YOLO格式的文本文件，该文件包含类别标签与边界框坐标等标注内容。

### 数据集标签
数据集所使用的类别标签如下：

['abiotic', 'insect', 'disease']

其中`abiotic`指代非生物胁迫，`insect`指代虫害，`disease`指代病害。

### 图像数量统计
数据集的图像分布如下：
json
{"train": 3788, "valid": 710, "test": 238}

训练集包含3788张图像，验证集包含710张图像，测试集包含238张图像。

### 标注实例数量统计
各分类的标注实例数量如下：
json
{"insect": 1618, "abiotic": 13960, "disease": 7032}

其中虫害实例1618个，非生物胁迫实例13960个，病害实例7032个。

### 解压后的文件夹结构
解压后的完整文件夹结构如下：
markdown
Data/
└── train/
├── images
├── labels
└── val/
├── images
├── labels
└── test/
├── images
├── labels

其中`train`为训练集文件夹，`val`为验证集文件夹，`test`为测试集文件夹，每个分类文件夹下均包含存储原始图像的`images`子文件夹与存储标注文件的`labels`子文件夹。

### 数据集详情
本数据集由卡拉Agro AI基金会（KaraAgro AI Foundation）的数据科学团队牵头创建，并获得了农业科学家与农技人员的技术支持。本数据集的制作得到了德国国际合作机构（Deutsche Gesellschaft für Internationale Zusammenarbeit, GIZ）的项目资助，包括其在西非国家经济共同体（ECOWAS）区域实施的《面向就业与增长的市场导向型价值链项目（Market-Oriented Value Chains for Jobs & Growth in the ECOWAS Region, MOVE）》，以及由GIZ代表德国联邦经济合作与发展部（Bundesministerium für Wirtschaftliche Zusammenarbeit und Entwicklung, BMZ）实施的「FAIR Forward——人工智能惠及人人」项目。

如需了解本数据集的详细信息，敬请查阅配套提供的数据集手册[此处](https://drive.google.com/file/d/1viv-PtZC_j9S_K1mPl4R1lFRKxoFlR_M/view?usp=sharing)。该手册详细阐述了数据集的构成、变量设置、数据采集方法及其他相关细节。