---
library_name: keras
license: apache-2.0
tags:
- potato-disease-classification
- plant-disease
- computer-vision
- agriculture
- densenet169
- image-classification
datasets:
- mendeley
model-index:
- name: densenet169_mendeley
  results:
  - task:
      type: image-classification
    metrics:
    - type: accuracy
      value: 0.9156
    - type: f1-score
      value: 0.9162
    - type: precision
      value: 0.9198
    - type: recall
      value: 0.9156
language:
- en
---

# DenseNet169 Potato Leaf Disease Classifier (Mendeley Dataset)

## Model Description

This model is a fine-tuned **DenseNet169** CNN trained for potato leaf disease classification on the **Mendeley Potato Disease Dataset**. The model is designed to detect and classify 7 different potato plant disease categories using leaf images.

### Key Features

- **Architecture**: DenseNet169 (pre-trained on ImageNet, fine-tuned on Mendeley dataset)
- **Framework**: TensorFlow/Keras
- **Input**: RGB images (224×224 pixels)
- **Classes**: 7 disease categories
- **Format**: `.keras` (Keras 3 format)

## Model Performance

### Overall Metrics
- **Accuracy**: 91.56%
- **Precision**: 91.98%
- **Recall**: 91.58%
- **F1-Score**: 91.62%
- **Matthews Correlation Coefficient (MCC)**: 0.8976
- **Balanced Accuracy**: 90.47%

### Per-Class Performance

| Class | Precision | Recall | F1-Score | Support |
|-------|-----------|--------|----------|---------|
| Bacteria | 0.97 | 0.98 | 0.98 | 114 |
| Fungi | 0.90 | 0.92 | 0.91 | 150 |
| Healthy | 0.75 | 0.97 | 0.85 | 40 |
| Nematode | 0.92 | 0.79 | 0.85 | 14 |
| Pest | 0.88 | 0.87 | 0.88 | 122 |
| Phytophthora | 0.94 | 0.91 | 0.93 | 69 |
| Virus | 0.98 | 0.89 | 0.93 | 107 |

**Test Set Size**: 616 images

## Dataset Information

### Mendeley Potato Disease Dataset
- **Total Images**: ~3,076 images
- **Classes**: 7 disease categories
  - Bacteria
  - Fungi
  - Healthy
  - Nematode
  - Pest
  - Phytophthora
  - Virus
- **Data Source**: [Mendeley Data - Potato Leaf Diseases (ptz377bwb8)](https://data.mendeley.com/datasets/ptz377bwb8/1)
- **Characteristics**: Real-world field images with varying lighting, angles, and natural conditions

## Usage

### Installation

```bash
pip install tensorflow keras
```

### Load and Use the Model

```python
import tensorflow as tf
from tensorflow import keras
from PIL import Image
import numpy as np

# Load the model
model = keras.models.load_model('best.keras')

# Prepare image
image = Image.open('path/to/potato_leaf.jpg')
image = image.resize((224, 224))
image_array = np.array(image) / 255.0
image_array = np.expand_dims(image_array, axis=0)

# Make prediction
predictions = model.predict(image_array)
class_names = ['Bacteria', 'Fungi', 'Healthy', 'Nematode', 'Pest', 'Phytophthora', 'Virus']
predicted_class = class_names[np.argmax(predictions)]
confidence = np.max(predictions)

print(f"Predicted Disease: {predicted_class}")
print(f"Confidence: {confidence:.2%}")
```

### Using Hugging Face Transformers

```python
# Coming soon - optimized inference pipeline
```

## Training Details

### Architecture
- **Base Model**: DenseNet169 (ImageNet pre-trained weights)
- **Input Size**: 224×224×3
- **Total Parameters**: ~14M
- **Custom Head**: Fine-tuned on Mendeley dataset

### Data Preprocessing
- Image resizing: 224×224
- Normalization: ImageNet normalization
- Augmentation: Random rotations, flips, brightness adjustments

### Training Configuration
- **Optimizer**: Adam
- **Loss Function**: Categorical Cross-Entropy
- **Metrics**: Accuracy, Precision, Recall, F1-Score
- **Early Stopping**: Enabled to prevent overfitting

## Model Card Information

- **Model Name**: DenseNet169 Potato Leaf Disease Classifier
- **Created**: 2026
- **Framework**: Keras/TensorFlow
- **Model Type**: Image Classification (CNN)
- **License**: Apache 2.0

## Intended Use

This model is intended for:
- **Primary Use**: Classification of potato leaf diseases in agricultural applications
- **Supported Use Cases**: 
  - Automated disease detection systems
  - Agricultural advisory systems
  - Precision farming applications
  - Research and baseline comparisons

## Limitations

- **Dataset Scope**: Trained exclusively on Mendeley dataset; performance may vary on other potato disease datasets
- **Class Imbalance**: Some classes (Nematode: 14 samples) have limited training data
- **Environmental Factors**: Model trained on field conditions; controlled environment images may perform differently
- **Image Quality**: Requires clear leaf images; heavily obscured or blurry images may reduce accuracy
- **Biological Variation**: Limited to potato plant diseases; not suitable for other crop diseases

## Related Models

- **DenseNet201 Mendeley**: Alternative architecture with similar performance
- **ResNet50 Mendeley**: Baseline comparison model
- **Models on PlantVillage Dataset**: 3-class variant (Early Blight, Late Blight, Healthy)

## Citation

If you use this model, please cite:

```bibtex
@dataset{mendeley_potato_disease,
  title = {Potato Leaf Diseases Dataset},
  url = {https://data.mendeley.com/datasets/ptz377bwb8/1}
}
```

## Contact & Support

For questions or issues regarding this model:
- Repository: [Potato-leaf-disease-prediction](https://github.com/krishna/Potato-leaf-disease-prediction)
- Dataset Issues: Reference the Mendeley dataset documentation

## Disclaimer

This model should be used as a reference tool and not as the sole basis for agricultural decisions. Always consult with agricultural experts for accurate disease diagnosis and treatment recommendations.

---

**Model Version**: 1.0  
**Last Updated**: May 2026  
**Framework**: TensorFlow 2.x / Keras 3  
**Python Version**: 3.8+