|
|
--- |
|
|
license: apache-2.0 |
|
|
tags: |
|
|
- computer-vision |
|
|
- image-segmentation |
|
|
- leaf-disease |
|
|
- agriculture |
|
|
--- |
|
|
|
|
|
# Annotated_Benchmarks400 |
|
|
|
|
|
This is a comprehensive collection of annotated "In-The-Lab" Tomato Leaf Images. Each Image has only one single leaf collected from the train split of Tomato subset from the PlanVillage Dataset from the Kaggle. |
|
|
The Precise Annotations are done using **Segment Anything 2.0** model with CVAT. The dataset contains Binary segmentation masks for each image. |
|
|
|
|
|
## Dataset Description |
|
|
|
|
|
This dataset is a combination of four individual leaf disease datasets, intended for training robust computer vision models for segmentation tasks in agriculture. It includes images of tomato leaves affected by various common diseases. |
|
|
|
|
|
The original datasets included are: |
|
|
- Septoria Leaf Spot |
|
|
- Yellow Leaf Curl Virus |
|
|
- Early Blight |
|
|
- Late Blight |
|
|
|
|
|
Here's the number of samples from each class: |
|
|
|
|
|
 |
|
|
|
|
|
### Features |
|
|
|
|
|
The dataset contains the following features for each example: |
|
|
|
|
|
- `image`: The original RGB image of the plant leaf. |
|
|
- `mask`: The binary, single-channel (grayscale) segmentation mask where the leaf/disease area is highlighted. |
|
|
- `image_id`: The original filename of the image, which can be used for traceability. |
|
|
- `width`: The original width of the image in pixels. |
|
|
- `height`: The original height of the image in pixels. |
|
|
- `num_annotations`: The number of individual polygons annotated in the image. |
|
|
|
|
|
## Dataset Structure |
|
|
|
|
|
The dataset consists of a single split ('train') with a total of **401** image-mask pairs. |
|
|
|
