Mango_Leaf_Bd_dataset

The Mango Leaf Disease Dataset has been systematically developed to facilitate research in plant disease detection and classification using deep learning and image processing techniques. The dataset focuses on mango leaves, an essential part of agricultural production in tropical and subtropical regions, and includes both healthy samples and leaves affected by common mango diseases.
1. Classes of the Dataset
The dataset is organized into four distinct classes:
1. Healthy
2. Anthracnose
3. Gall Midges
4. Powdery Mildew
2. Dataset Composition
• Original Images:
o 500 images per class (total 2,000 original samples).
o Image format: JPG.
• Augmented Images:
o To enhance variability and overcome data scarcity, augmentation techniques such as rotation, flipping, scaling, brightness adjustment, contrast variation, and Gaussian noise addition were applied.
o 512 × 512 resolution: 1,500 images per class.
o 256 × 256 resolution: 1,500 images per class.
o Image format: JPG.
3. Final Dataset Size
• 512 × 512 images: 6,000 samples (1,500 × 4 classes).
• 256 × 256 images: 6,000 samples (1,500 × 4 classes).
• Grand Total (original + augmented): 14,000 JPG images.