import streamlit as st
import cv2
import numpy as np
import matplotlib.pyplot as plt
from sklearn.cluster import KMeans
import matplotlib.ticker as ticker
from io import BytesIO
from PIL import Image

# -----------------------------
# Streamlit App
# -----------------------------
st.title("🖼️ Image Segmentation using K-Means Clustering")

# Sidebar options
st.sidebar.header("Controls")
uploaded_file = st.sidebar.file_uploader("Upload an image", type=["jpg", "jpeg", "png"])
max_k = st.sidebar.slider("Max K for Elbow Curve", 5, 15, 10)
chosen_k = st.sidebar.slider("Choose K (Clusters)", 2, 10, 4)

if uploaded_file is not None:
    # -----------------------------
    # Step 1: Load Image
    # -----------------------------
    image = Image.open(uploaded_file)
    image = np.array(image)

    st.subheader("Original Image")
    st.image(image, use_container_width=True)

    # -----------------------------
    # Step 2: Preprocess Image
    # -----------------------------
    pixels = image.reshape((-1, 3))
    pixels = np.float32(pixels)

    # -----------------------------
    # Step 3: Elbow Curve
    # -----------------------------
    wcss = []
    K = range(1, max_k+1)
    for k in K:
        kmeans = KMeans(n_clusters=k, random_state=42)
        kmeans.fit(pixels)
        wcss.append(kmeans.inertia_)

    # Plot elbow curve
    fig, ax = plt.subplots(figsize=(6, 4))
    ax.plot(K, wcss, 'bo-')
    ax.set_xlabel("Number of clusters (k)")
    ax.set_ylabel("WCSS (Inertia)")
    ax.set_title("Elbow Curve")

    # Format y-axis in millions
    ax.yaxis.set_major_formatter(ticker.FuncFormatter(lambda x, _: f'{int(x/1e6)}M'))

    # Add vertical dotted line for chosen k
    ax.axvline(x=chosen_k, color='r', linestyle='--', label=f'k={chosen_k}')
    ax.legend()

    st.subheader("Elbow Curve")
    st.pyplot(fig)

    # -----------------------------
    # Step 4: Apply KMeans with chosen k
    # -----------------------------
    kmeans = KMeans(n_clusters=chosen_k, random_state=42)
    labels = kmeans.fit_predict(pixels)

    segmented_img = kmeans.cluster_centers_[labels]
    segmented_img = segmented_img.reshape(image.shape)
    segmented_img = np.uint8(segmented_img)

    # -----------------------------
    # Step 5: Show Results
    # -----------------------------
    st.subheader(f"Segmented Image (k={chosen_k})")
    st.image(segmented_img, use_container_width=True)

else:
    st.info("👈 Upload an image from the sidebar to start.")