File size: 4,264 Bytes
d63d884
 
74df710
 
 
 
 
 
 
 
 
 
 
 
d63d884
74df710
 
 
a33bf60
74df710
a33bf60
85581ba
a5818b5
 
 
85581ba
a33bf60
a5818b5
a33bf60
 
 
 
 
fc7e9e0
a33bf60
 
 
 
 
a5818b5
a33bf60
 
fc7e9e0
 
 
 
 
 
 
a33bf60
a5818b5
85581ba
a33bf60
 
 
 
 
 
 
 
 
 
 
 
85581ba
a33bf60
fba478c
a33bf60
 
 
 
 
 
 
 
 
fba478c
a33bf60
74df710
a33bf60
74df710
3a19e12
a33bf60
 
3a19e12
a33bf60
3a19e12
 
74df710
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
---
license: cc-by-4.0
datasets:
- KaraAgroAI/KaraAgroAI/Drone-based-Agricultural-Dataset-for-Crop-Yield-Estimation
language:
- en
library_name: yolo
tags:
- object detection
- vision
- yolo
pipeline_tag: object-detection
metrics:
- mape
---

## Drone-based Agricultural Dataset for Crop Yield Estimation

This repository contains a comprehensive dataset of cashew, cocoa and coffee images captured by drones, accompanied by meticulously annotated labels. To facilitate 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.

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)

## Motivation
This dataset bridges a gap by offering a comprehensive collection of agricultural images specifically designed to fuel the development and evaluation of yield estimation models. Estimating crop yield accurately is a complex task influenced by numerous factors including weather conditions, soil quality, pest prevalence, and cultivation practices. By offering a diverse range of images capturing different crops, growth stages, and environmental conditions, this dataset empowers researchers, data scientists, and agronomists to develop models that are robust and adaptable to the variability inherent in real-world agricultural scenarios.


### Ghana - KaraAgro AI
Each image in the Ghana set has a resolution of 16000 by 13000 pixels. There is a total of 8,784 images and annotations in the Ghana set.

#### Dataset Labels

```
Cashew --> ['cashew_tree', 'flower', 'immature', 'mature', 'ripe', 'spoilt']
Cocoa --> ['cocoa-pod-mature-unripe', 'cocoa-tree', 'cocoa-pod-immature', 'cocoa-pod-riped', 'cocoa-pod-spoilt']
```

#### Number of Images
```json
Cashew --> 4,715 images
Cocoa --> 4,069 images
```

### Number of Instances Annotated

```json
Cashew --> {'cashew_tree':1107, 'flower':16757, 'immature':11766, 'mature': 4244, 'ripe': 11721, 'spoilt': 518}
Cocoa --> {'cocoa-pod-mature-unripe': 10786, 'cocoa-tree': 2831, 'cocoa-pod-immature': 2401, 'cocoa-pod-riped': 4193, 'cocoa-pod-spoilt': 2018}
```

### Uganda
A total of 6,086 drone images, comprising 3,000 for coffee and 3,086 for cashew. Each image in the Uganda set has dimensions of 4,000 by 3,000 pixels.

#### Dataset Labels

```
Cashew --> ['cashew_tree', 'flower', 'immature', 'mature', 'ripe', 'spoilt']
Coffee --> ['unripe', 'ripening', 'ripe', 'spoilt', 'coffee']
```

#### Number of Images
```json
Cashew --> 3,086 images
Coffee --> 3,000 images
```

### Folder structure

```markdown
Data/
└── Ghana/
    β”œβ”€β”€ cashew.zip
    β”œβ”€β”€ cocoa.zip
└── Uganda/
    β”œβ”€β”€ cashew.zip
    β”œβ”€β”€ coffee.zip
```

### Intended uses

The dataset which was mainly developed for yield estimation can also be usedfor further research including crop abnormality detection due to the presence of spoilt classes in the datasets


### Dataset Information
The dataset was created by a team of data scientists from the KaraAgro AI Foundation, with support from the agricultural scientists and officers. The creation of this dataset was made possible through the funding from the Lacuna Fund. For detailed information regarding the datasheet, we invite you to explore the accompanying datasheet available [here](https://). This comprehensive resource offers a deeper understanding of the dataset's compostion, variables, data collection methodologies, and othe relevant details.

Our datasets comply with the Findable, Accessible, InterOperable, and Reusable (FAIR) data principles. The datasets have been assigned a Digital Object Identifier (DOI), a permanent unique identifier to facilitate findability and accessibility.  The metadata is citable and includes domain-specific and file-level data that map to metadata standards within machine learning, computer vision, data analysis - geospatial and time series analysis to make it Interoperable. Our datasets along with their associated metadata may be accessed and downloaded via this link : <a href = "https://doi.org/10.57967/hf/0959"> doi.org/10.57967/hf/0959 </a>