import cv2
import face_recognition
import numpy as np
import gradio as gr

# Load reference image and encode face
reference_image = face_recognition.load_image_file("reference.jpg")
reference_encoding = face_recognition.face_encodings(reference_image)[0]

def detect_and_recognize(image_path):
    # Load input image
    image = face_recognition.load_image_file(image_path)
    
    # Detect faces
    face_locations = face_recognition.face_locations(image)
    face_encodings = face_recognition.face_encodings(image, face_locations)

    # Convert image for OpenCV visualization
    image_cv2 = cv2.imread(image_path)
    
    recognized_faces = []

    for (top, right, bottom, left), face_encoding in zip(face_locations, face_encodings):
        # Compare detected face with reference
        matches = face_recognition.compare_faces([reference_encoding], face_encoding)
        label = "Match" if matches[0] else "Unknown"
        
        # Draw a rectangle around the face
        cv2.rectangle(image_cv2, (left, top), (right, bottom), (0, 255, 0), 2)
        cv2.putText(image_cv2, label, (left, top - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)

        recognized_faces.append(label)
    
    output_path = "output.jpg"
    cv2.imwrite(output_path, image_cv2)

    return output_path, recognized_faces

# Gradio Interface
iface = gr.Interface(fn=detect_and_recognize, inputs="file", outputs=["image", "text"], title="Face Recognition")
iface.launch()