import streamlit as st
from streamlit_webrtc import webrtc_streamer, VideoTransformerBase
import mediapipe as mp
import cv2
import av

# Initialize MediaPipe Hands
mp_hands = mp.solutions.hands
mp_drawing = mp.solutions.drawing_utils

class VideoProcessor(VideoTransformerBase):
    def __init__(self):
        # Initialize MediaPipe Hands
        self.hands = mp_hands.Hands(
            max_num_hands=2,
            min_detection_confidence=0.5,
            min_tracking_confidence=0.5,
        )

    def recv(self, frame):
        # Convert the input frame to a numpy array
        img = frame.to_ndarray(format="bgr24")
        img = cv2.flip(img, 1)  # Flip horizontally for a mirror effect
        img_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

        # Process the frame for hand landmarks
        result = self.hands.process(img_rgb)

        # Draw hand landmarks on the frame
        if result.multi_hand_landmarks:
            for hand_landmarks in result.multi_hand_landmarks:
                mp_drawing.draw_landmarks(
                    img, hand_landmarks, mp_hands.HAND_CONNECTIONS
                )

        # Return the processed frame back to the WebRTC streamer
        return av.VideoFrame.from_ndarray(img, format="bgr24")

# Streamlit UI
st.title("Gesture & Hand Landmark Detection 🚀")
st.write("This app uses MediaPipe and Streamlit to detect hand landmarks in real-time from your webcam.")

# WebRTC streamer for live video feed
webrtc_streamer(
    key="gesture-detection",
    video_processor_factory=VideoProcessor,
    rtc_configuration={
        "iceServers": [{"urls": ["stun:stun.l.google.com:19302"]}]
    },
)

# Footer
st.markdown(
    """
    <style>
    .footer {text-align: center; font-size: 12px; color: grey; margin-top: 20px;}
    </style>
    <p class="footer">Made with g Streamlit & MediaPipe</p>
    """,
    unsafe_allow_html=True,
)