main.py

import os
import json
from PIL import Image
import numpy as np
import tensorflow as tf
import streamlit as st
import gdown

# Define the working directory and paths
working_dir = os.path.dirname(os.path.abspath(__file__))
model_path = os.path.join(working_dir, "model.h5")
class_indices_path = os.path.join(working_dir, "class_indices.json")

# URL of the model on Google Drive (Direct download link)
model_url = "https://drive.google.com/uc?id=1cgOXZQkK1Yt60b0Ry0wUZoc2fSaQhW_V"

# Download the model if it doesn't exist
if not os.path.exists(model_path):
    gdown.download(model_url, model_path, quiet=False)

# Load the pre-trained model
model = tf.keras.models.load_model(model_path)

# Load the class names
with open(class_indices_path, "r") as f:
    class_indices = json.load(f)

# Function to Load and Preprocess the Image using Pillow
def load_and_preprocess_image(image, target_size=(224, 224)):
    img = Image.open(image)
    img = img.resize(target_size)
    img_array = np.array(img)
    img_array = np.expand_dims(img_array, axis=0)
    img_array = img_array.astype('float32') / 255.
    return img_array

# Function to Predict the Class of an Image
def predict_image_class(model, image, class_indices):
    preprocessed_img = load_and_preprocess_image(image)
    predictions = model.predict(preprocessed_img)
    predicted_class_index = np.argmax(predictions, axis=1)[0]
    predicted_class_name = class_indices[str(predicted_class_index)]
    return predicted_class_name

# Streamlit App
st.header("Plant Disease Recognition System 🌿🔍")

# Path to the image to be displayed on the homepage
home_image_path = os.path.join(working_dir, "home_page.jpg")
st.image(home_image_path, use_column_width=True)

st.markdown("""
### Introduction
Welcome to the Plant Disease Recognition System, a cutting-edge project developed for our final year project. Our system aims to assist in the efficient identification of plant diseases. By uploading an image of a plant, our system can quickly analyze it to detect any signs of diseases, helping to protect crops and ensure a healthier harvest.
     
### Team Members
- Aditi Tomar (Roll No. - 200060101014)
- Ansh Agarwal (Roll No. - 200060101026)
- Atul Pratap Singh (Roll No. - 200060101046)
- Ankit Bisht (Roll No. - 200060101025) 
    
### How It Works
- **Upload Image:** Navigate to the **Prediction** page and upload an image of a plant showing symptoms of disease.
- **Analysis:** Our system utilizes state-of-the-art machine learning algorithms to process the image and identify potential diseases.
- **Results:** View detailed results and receive recommendations for further action.

### Features
- **Accuracy:** Our system employs advanced machine learning techniques for accurate disease detection.
- **User-Friendly:** Enjoy a simple and intuitive interface designed for a seamless user experience.
- **Fast and Efficient:** Receive results in seconds, enabling quick decision-making in the field.

### Dataset    
This dataset is part of a crowdsourcing effort to leverage smartphone technology and machine learning for improving food production by addressing the issue of infectious diseases in crops.
- Size: Over 50,000 expertly curated images.
- Content: The images consist of healthy and infected leaves of crop plants.
- Platform: The dataset is released on the PlantVillage online platform.
- Purpose: The dataset is intended to facilitate the development of mobile disease diagnostics using machine learning and crowdsourcing.
- Goal: The goal is to use computer vision approaches to address yield losses in crop plants caused by infectious diseases.

### Prediction
Upload an image and experience the power of our Plant Disease Recognition System!
""")

uploaded_image = st.file_uploader("Upload an image", type=["jpg", "jpeg", "png"], label_visibility="hidden")

if uploaded_image is not None:
    image = Image.open(uploaded_image)
    col1, col2 = st.columns(2)

    with col1:
        resized_img = image.resize((150, 150))
        st.image(resized_img)

    with col2:
        if st.button('Classify'):
            prediction = predict_image_class(model, uploaded_image, class_indices)
            st.success(f'Prediction: {prediction}')