Yadvendra/Brain_Tumor_Detection_Using_Tensorflow

1import streamlit as st2import numpy as np3import cv24import tensorflow as tf5from PIL import Image6from sklearn.preprocessing import LabelEncoder7 8# Load your pre-trained model (Make sure this matches the version used during training)9model = tf.keras.models.load_model('brain_tumor_model.h5')10 11# Example class labels (update this list with your actual class labels)12class_labels = ['glioma', 'pituitary', 'meningioma', 'healthy']13label_encoder = LabelEncoder()14label_encoder.fit(class_labels)  # Fit the label encoder with your class labels15 16# Function to load and preprocess the uploaded image17def load_and_preprocess_image(uploaded_file):18    img = Image.open(uploaded_file)19    img = img.convert("RGB")  # Convert to RGB if it's in another format20    img = np.array(img)  # Convert to NumPy array21    img = cv2.resize(img, (224, 224))  # Resize the image to 224x22422    img = img / 255.0  # Normalize pixel values23    img = np.reshape(img, (1, 224, 224, 3))  # Reshape for prediction24    return img25 26# Function to predict the image class27def predict_image(img):28    predictions = model.predict(img)  # Make a prediction29    predicted_class_index = np.argmax(predictions[0])  # Get the predicted class index30    return predicted_class_index31 32# Function to get class label33def get_class_label(predicted_class_index):34    return label_encoder.inverse_transform([predicted_class_index])[0]  # Get class label35 36# Streamlit App UI37st.title("Brain Tumor using CNN 🧠")38st.write("Upload a brain scan (JPG format), and the model will predict its class.")39 40# File uploader for user to upload images41uploaded_file = st.file_uploader("Choose a JPG image...", type="jpg")42 43if uploaded_file is not None:44    # Display the uploaded image on the left side45    col1, col2 = st.columns([2, 1])  # Create two columns46 47    with col1:48        st.image(uploaded_file, caption="Uploaded Image", use_column_width=True)49    50    with col2:51        # Button to trigger prediction52        if st.button("Detect"):53            st.write("Detecting...")54            # Load and preprocess the image55            processed_image = load_and_preprocess_image(uploaded_file)56            57            # Make prediction58            predicted_class_index = predict_image(processed_image)59            60            # Get predicted class label61            predicted_class_label = get_class_label(predicted_class_index)62            63            # Center display for the prediction result64            st.markdown(f"<h3 style='color: #4CAF50; text-align: center;'>The Prediction is : <strong>{predicted_class_label}</strong></h3>", unsafe_allow_html=True)65 66