app.py

import streamlit as st
import tensorflow as tf
from tensorflow import keras
import numpy as np
from PIL import Image
import cv2
import matplotlib.pyplot as plt

# Page configuration
st.set_page_config(
    page_title="Microscopic Organism Classifier",
    page_icon="🔬",
    layout="wide"
)

# Custom CSS
st.markdown("""
    <style>
    .main-header {
        font-size: 42px;
        font-weight: bold;
        color: #2E86AB;
        text-align: center;
        margin-bottom: 30px;
    }
    .sub-header {
        font-size: 20px;
        color: #666;
        text-align: center;
        margin-bottom: 40px;
    }
    .prediction-box {
        background-color: #f0f8ff;
        padding: 20px;
        border-radius: 10px;
        border-left: 5px solid #2E86AB;
        margin: 20px 0;
    }
    </style>
""", unsafe_allow_html=True)

# Load model
@st.cache_resource
def load_model():
    try:
        model = keras.models.load_model('microscopic_organism_classifier.h5')
        return model
    except:
        st.warning("Model file not found. Please train the model first.")
        return None

# Class names (update based on your dataset)
CLASS_NAMES = ['Amoeba', 'Diatom', 'Euglena', 'Paramecium', 'Volvox']

def preprocess_image(image, target_size=(128, 128)):
    """Preprocess uploaded image"""
    img = image.resize(target_size)
    img_array = np.array(img) / 255.0
    img_array = np.expand_dims(img_array, axis=0)
    return img_array

def predict_organism(model, image):
    """Make prediction on image"""
    processed_img = preprocess_image(image)
    predictions = model.predict(processed_img)[0]
    predicted_class_idx = np.argmax(predictions)
    predicted_class = CLASS_NAMES[predicted_class_idx]
    confidence = predictions[predicted_class_idx] * 100
    return predicted_class, confidence, predictions

# Main UI
st.markdown('<p class="main-header">🔬 Microscopic Organism Classifier</p>', unsafe_allow_html=True)
st.markdown('<p class="sub-header">Upload microscopic images for instant AI-powered identification</p>', unsafe_allow_html=True)

# Sidebar
with st.sidebar:
    st.header("📋 About")
    st.info("""
    This application uses deep learning (CNN) to classify microscopic organisms from images.
    
    **Supported Organisms:**
    - Amoeba
    - Diatom
    - Euglena
    - Paramecium
    - Volvox
    
    **How to use:**
    1. Upload an image
    2. Get instant classification
    3. View confidence scores
    """)
    
    st.header("⚙️ Model Info")
    st.metric("Model Type", "CNN")
    st.metric("Input Size", "128x128")
    st.metric("Classes", len(CLASS_NAMES))

# Main content
col1, col2 = st.columns([1, 1])

with col1:
    st.subheader("📤 Upload Image")
    uploaded_file = st.file_uploader(
        "Choose a microscopic organism image",
        type=['png', 'jpg', 'jpeg'],
        help="Upload a clear microscopic image for best results"
    )
    
    if uploaded_file is not None:
        image = Image.open(uploaded_file)
        st.image(image, caption='Uploaded Image', use_container_width=True)
        
        # Classify button
        if st.button('🔍 Classify Organism', type='primary', use_container_width=True):
            model = load_model()
            
            if model is not None:
                with st.spinner('Analyzing image...'):
                    predicted_class, confidence, all_predictions = predict_organism(model, image)
                
                with col2:
                    st.subheader("📊 Results")
                    
                    # Prediction box
                    st.markdown(f"""
                    <div class="prediction-box">
                        <h2 style="color: #2E86AB; margin: 0;">Predicted: {predicted_class}</h2>
                        <h3 style="color: #666; margin-top: 10px;">Confidence: {confidence:.2f}%</h3>
                    </div>
                    """, unsafe_allow_html=True)
                    
                    # Probability chart
                    st.subheader("Probability Distribution")
                    fig, ax = plt.subplots(figsize=(10, 6))
                    colors = ['#2E86AB' if i == np.argmax(all_predictions) else '#A7C6D9' 
                             for i in range(len(CLASS_NAMES))]
                    ax.barh(CLASS_NAMES, all_predictions, color=colors)
                    ax.set_xlabel('Probability', fontsize=12)
                    ax.set_xlim([0, 1])
                    ax.grid(axis='x', alpha=0.3)
                    plt.tight_layout()
                    st.pyplot(fig)
                    
                    # Detailed probabilities
                    st.subheader("Detailed Probabilities")
                    for i, class_name in enumerate(CLASS_NAMES):
                        prob = all_predictions[i] * 100
                        st.progress(all_predictions[i])
                        st.text(f"{class_name}: {prob:.2f}%")
                    
                    # Interpretation
                    if confidence > 90:
                        st.success("✅ High confidence prediction!")
                    elif confidence > 70:
                        st.info("ℹ️ Moderate confidence. Consider image quality.")
                    else:
                        st.warning("⚠️ Low confidence. Please upload a clearer image.")

# Sample images section
st.markdown("---")
st.subheader("📸 Sample Test Images")
st.info("Don't have an image? Download sample microscopic organism images from these sources: [Kaggle Plankton Dataset](https://www.kaggle.com) or Google Images")

# Footer
st.markdown("---")
st.markdown("""
<div style="text-align: center; color: #666; padding: 20px;">
    <p>Developed with ❤️ using TensorFlow & Streamlit</p>
    <p>For production deployment, contact me for full implementation with 20+ organism classes</p>
</div>
""", unsafe_allow_html=True)