Tomato Disease Detection API

A Flask REST API for detecting tomato leaf diseases from images using a fine-tuned MobileViT model. The model classifies tomato leaf images into four categories: Early Blight, Late Blight, Septoria Leaf Spot, and Healthy.

Model

The API uses a MobileViT-S model (mobilevit_s_tomato.pth) fine-tuned on tomato disease data, with image preprocessing via Apple's mobilevitv2-1.0-imagenet1k-256 processor.

Supported classes:

Label	Description
early blight	Caused by Alternaria solani fungus
late blight	Caused by Phytophthora infestans
septoria leaf spot	Caused by Septoria lycopersici fungus
healthy	No disease detected

Requirements

• Python 3.11+
• See requirements.txt for full dependency list

Setup

# Create and activate virtual environment
python3 -m venv venv
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

Create a .env file in the project root:

SECRET_KEY=your_secret_key
JWT_SECRET_KEY=your_jwt_secret_key
FLASK_DEBUG=True

Running the API

flask --app app run

The API will be available at http://127.0.0.1:5000.
Interactive Swagger UI: http://127.0.0.1:5000/

Database Migrations

flask db init      # first time only
flask db migrate
flask db upgrade

---

Authentication

All classification endpoints require a JWT access token. Include it in the Authorization header:

Authorization: Bearer <access_token>

Access tokens expire after 5 minutes. Use the refresh token (valid for 20 minutes) to obtain a new access token.

---

API Reference

Auth

POST /auth/register

Create a new user account.

Request body:

{
  "firstname": "Jane",
  "lastname": "Doe",
  "username": "janedoe",
  "password": "secret"
}

Responses:

Status	Description
201 Created	User created successfully
409 Conflict	Username already exists

---

POST /auth/login

Authenticate and receive JWT tokens.

Request body:

{
  "username": "janedoe",
  "password": "secret"
}

Response 200 OK:

{
  "access_token": "<jwt_access_token>",
  "refresh_token": "<jwt_refresh_token>",
  "user_name": "janedoe",
  "id": 1
}

Status	Description
200 OK	Login successful
401 Unauthorized	Invalid credentials

---

GET /auth/refresh

Obtain a new access token using a refresh token.

Headers:

Authorization: Bearer <refresh_token>

Response 200 OK:

{
  "access_token": "<new_jwt_access_token>"
}

---

Classification

POST /classification/classify

Upload a tomato leaf image for disease classification. Requires a valid access token.

Headers:

Authorization: Bearer <access_token>
Content-Type: multipart/form-data

Form data:

Field	Type	Description
image	file	Tomato leaf image (.jpg, .jpeg, or .png)

Response 200 OK:

{
  "label": "early blight",
  "image_url": "http://127.0.0.1:5000/api/uploads/1703704042311672.jpg",
  "label_probabilities": {
    "Early Blight": 0.8731,
    "Late Blight": 0.0423,
    "Septoria Leaf Spot": 0.0612,
    "Healthy": 0.0234
  }
}

Status	Description
200 OK	Classification successful
400 Bad Request	No file provided or invalid file type
401 Unauthorized	Missing or invalid token
500 Internal Server Error	Model inference error

---

GET /classification/classification_results/<user_id>

Retrieve paginated classification history for a user. Requires a valid access token.

Headers:

Authorization: Bearer <access_token>

Query parameters:

Parameter	Default	Description
page	1	Page number
per_page	2	Results per page

Response 200 OK:

{
  "results": [
    {
      "image_url": "http://127.0.0.1:5000/api/uploads/1703704042311672.jpg",
      "result_value": "early blight",
      "timestamp": "2024-01-15T10:30:00"
    }
  ],
  "total_pages": 5
}

Status	Description
200 OK	Results returned
204 No Content	No results for this user
404 Not Found	User does not exist

---

GET /classification/image/<image_file_name>

Retrieve a previously uploaded image by filename. Does not require authentication.

Example:

GET /classification/image/1703704042311672.jpg

Status	Description
200 OK	Image returned
404 Not Found	Image not found

---

Project Structure

mobilevit-flask-api/
├── app.py                      # Application entry point
├── mobilevit_s_tomato.pth      # Fine-tuned MobileViT model
├── requirements.txt
├── .env
├── api/
│   ├── __init__.py             # App factory
│   ├── auth/
│   │   └── views.py            # Register, login, refresh endpoints
│   ├── classification/
│   │   └── views.py            # Classify, history, image endpoints
│   ├── config/
│   │   └── config.py           # Dev/prod/test config
│   ├── models/
│   │   ├── users.py
│   │   └── classification_results.py
│   ├── uploads/                # Stored uploaded images
│   └── utils/
│       └── __init__.py         # SQLAlchemy db instance
└── instance/
    └── db.sqlite3              # SQLite database