🐛 Pest Detection & Pesticide Recommendation System — Flutter + Flask + AI

This application is a full-stack AI pest management tool that detects common agricultural pests from images and recommends natural, non-harmful pesticides.
It combines a TensorFlow MobileNetV2 model for classification, a Flask REST API backend, and a Flutter mobile frontend for an end-to-end pest detection workflow.

The system accepts pest images, classifies them into one of 12 categories, fetches natural pesticide recommendations from a curated dataset, and displays results instantly in a mobile app.

---

📸 App Screenshots

-----

⭐ Core Features

• 🦋 Pest Image Classification: Detects pests such as ant, bee, beetle, caterpillar, earthworm, earwig, grasshopper, moth, slug, snail, wasp, and weevil.
• 🌱 Natural Pesticide Recommendation: Suggests eco-friendly, non-chemical pesticides from a local dataset.
• 📱 Mobile Frontend: Flutter-based UI for easy image uploads and instant results.
• ⚡ REST API Backend: Flask API processes uploaded images and returns pest classification + pesticide suggestion in JSON format.
• 📦 Offline Support: TFLite model version available for offline mobile predictions.

---

🛠️ How It Works

1. Image Capture/Upload: User takes or selects a pest image in the Flutter app.
2. API Request: Image is sent to the Flask backend via POST request.
3. Preprocessing: The image is resized to 224×224 and normalized using MobileNetV2’s preprocess function.
4. Prediction: The TensorFlow model outputs class probabilities; the highest probability is chosen as the detected pest.
5. Recommendation: The pest name is matched in a CSV dataset to retrieve the recommended natural pesticide.
6. Response: Backend sends a JSON response with the pest name and recommended pesticide, displayed in the app.

---

🚀 Technology Stack

• Backend: Python, Flask, TensorFlow, Pandas, Pillow
• Frontend: Flutter, Dart
• Model: MobileNetV2 (fine-tuned)
• Data Storage: CSV for pesticide mapping
• Deployment: Local or cloud-hosted Flask API

---

🔧 Setup and Installation

1. Prerequisites

• Python 3.9+
• Flutter SDK installed
• pip for Python dependencies

2. Clone the Repository

git clone <repository-url>
cd <repository-directory>

3. Backend Setup

cd backend
pip install -r requirements.txt
python app.py

4. Frontend Setup

cd frontend_flutter
flutter pub get
flutter run

---

▶️ Usage

Backend API Endpoint:

POST /predict

Form Data:

• image — pest image file

Example:

curl -X POST -F "image=@ant.jpg" http://localhost:5000/predict

Response:

{
  "pest": "ant",
  "recommended_pesticide": "Neem Oil"
}

---

📂 Project Structure

project/
├── backend/
│   ├── app.py                 # Flask API
│   ├── data/                  # CSV pesticide dataset
│   ├── model/                 # Trained .h5 model
│   ├── uploads/               # Uploaded images
│   └── requirements.txt
│
├── frontend_flutter/          # Flutter app
│   ├── lib/                   # Dart code
│   ├── assets/
│   ├── android/ ios/ web/ etc.
│
├── ML/
│   ├── agriculture pests image dataset/
│   └── notebook/              # Jupyter notebook & models
│
├── image_test/                # Test images
├── demo/                      # App screenshots
└── README.md