📌Project Overview

Tomato crops are highly vulnerable to fungal diseases such as Early Blight, Late Blight, and Leaf Mold. Early detection and understanding of environmental risks can significantly reduce crop loss. This system provides:

• Automated disease classification using a CNN model

• Environmental risk scoring using simulated temperature/humidity data

• Actionable recommendations for farmers

• End-to-end pipeline + Streamlit interface

The combination of image classification + environmental risk assessment represents the novel contribution of this work.

🚀 Features

✔ CNN-Based Disease Detection

Detects 3 tomato diseases:

• Tomato Early Blight

• Tomato Late Blight

• Tomato Leaf Mold

Using an EfficientNet-based classifier.

✔ Environmental Risk Modeling

Simulated environmental data for:

• Pakistan

• California

• The Netherlands

• India

Model outputs:

• Region-specific disease spread likelihood (0 to 1)

✔ Recommendations Engine

Outputs:

• Short actionable advice

• Detailed expert recommendation

✔ Streamlit UI

User-friendly interface:

• Upload image

• View predicted disease

• View risk scores

• Get recommendations

📊 Dataset

• PlantVillage Tomato Subset (Kaggle)

• 3 classes:

o Tomato Early Blight

o Tomato Late Blight

o Tomato Leaf Mold

Environmental data is synthetically generated based on realistic agricultural ranges.

🔧 Tools & Technologies

• Python

• TensorFlow / Keras

• Pandas

• NumPy

• Streamlit

• Matplotlib