face_verification.py

import face_recognition
from deepface import DeepFace
import numpy as np
import cv2
import os
import logging
from utils import file_exists, read_yaml


logging_str = "[%(asctime)s: %(levelname)s: %(module)s]: %(message)s"
log_dir = "logs"
os.makedirs(log_dir, exist_ok=True)
logging.basicConfig(filename=os.path.join(log_dir,"ekyc_logs.log"), level=logging.INFO, format=logging_str, filemode="a")


config_path = "config.yaml"
config = read_yaml(config_path)

artifacts = config['artifacts']
cascade_path = artifacts['HAARCASCADE_PATH']
output_path = artifacts['INTERMIDEIATE_DIR']

def detect_and_extract_face(img):

    # Read the image
    # img = cv2.imread(image_path)

    # Convert the image to grayscale (Haar cascade works better with grayscale images)
    gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

    # Load the Haar cascade classifier
    face_cascade = cv2.CascadeClassifier(cascade_path)

    # Detect faces in the image
    faces = face_cascade.detectMultiScale(gray_img, scaleFactor=1.1, minNeighbors=4)
    # Find the face with the largest area
    max_area = 0
    largest_face = None
    for (x, y, w, h) in faces:
        area = w * h
        if area > max_area:
            max_area = area
            largest_face = (x, y, w, h)

    # Extract the largest face
    if largest_face is not None:
        (x, y, w, h) = largest_face
        # extracted_face = img[y:y+h, x:x+w]
        
        # Increase dimensions by 15%
        new_w = int(w * 1.50)
        new_h = int(h * 1.50)
        
        # Calculate new (x, y) coordinates to keep the center of the face the same
        new_x = max(0, x - int((new_w - w) / 2))
        new_y = max(0, y - int((new_h - h) / 2))

        # Extract the enlarged face
        extracted_face = gray_img[new_y:new_y+new_h, new_x:new_x+new_w]

        # Convert the extracted face to RGB
        # extracted_face_rgb = cv2.cvtColor(extracted_face, cv2.COLOR_BGR2RGB)

        
        current_wd = os.getcwd()
        filename = os.path.join(current_wd, output_path, "extracted_face.jpg")
        print(filename)
        print("*****************")

        if os.path.exists(filename):
            # Remove the existing file
            os.remove(filename)

        cv2.imwrite(filename, extracted_face)
        print(f"Extracted face saved at: {filename}")
        return filename

        # return extracted_face_rgb
    else:
        return None

def face_recog_face_comparison(image1_path="data\\02_intermediate_data\\extracted_face.jpg", image2_path = "data\\02_intermediate_data\\face_image.jpg"):

    img1_exists = file_exists(image1_path)
    img2_exists = file_exists(image2_path)

    if img1_exists and img2_exists:
        print("Check the path for the images provided")
        return False

    image1 = face_recognition.load_image_file(image1_path)
    image2 = face_recognition.load_image_file(image2_path)

    if image1 is not None and image2 is not None:
        face_encodings1 = face_recognition.face_encodings(image1)
        face_encodings2 = face_recognition.face_encodings(image2)

    else:
        print("Image is not loaded properly")
        return False

    # print(face_encodings1)

    # Check if faces are detected in both images
    if len(face_encodings1) == 0 or len(face_encodings2) == 0:
        print("No faces detected in one or both images.")
        return False
    else:
    # Proceed with comparing faces if faces are detected
        matches = face_recognition.compare_faces(np.array(face_encodings1), np.array(face_encodings2))
    # Print the results
    if matches[0]:
        print("Faces are verified")
        return True
    else:
        # print("The faces are not similar.")
        return False
    
def deepface_face_comparison(image1_path="data\\02_intermediate_data\\extracted_face.jpg", image2_path = "data\\02_intermediate_data\\face_image.jpg"):
    # img_path1 = "data\\01_raw_data\\bibek_face.jpg"
    # img_path2 = "data\\01_raw_data\\bibek_face2.jpg"
    # data\01_raw_data\bibek_face.jpg
    img1_exists = file_exists(image1_path)
    img2_exists = file_exists(image2_path)

    if not(img1_exists or img2_exists):
        print("Check the path for the images provided")
        return False
    
    verfication = DeepFace.verify(img1_path=image1_path, img2_path=image2_path)

    if len(verfication) > 0 and verfication['verified']:
        print("Faces are verified")
        return True
    else:
        return False

def face_comparison(image1_path, image2_path, model_name = 'deepface'):

    is_verified = False
    if model_name == 'deepface':
        is_verified = deepface_face_comparison(image1_path, image2_path)
    elif model_name ==  'facerecognition':
        is_verified = face_recog_face_comparison(image1_path, image2_path)
    else:
        print("Mention proper model name for face recognition")

    return is_verified

def get_face_embeddings(image_path):

    img_exists = file_exists(image_path)

    if not(img_exists):
        print("Check the path for the images provided")
        return None
    
    embedding_objs = DeepFace.represent(img_path = image_path, model_name = "Facenet")
    embedding = embedding_objs[0]["embedding"]

    if len(embedding) > 0:
        return embedding
    return None


if __name__ == "__main__":

    id_card = "data\\01_raw_data\\pan_2.jpg"
    # face_path = "data\\01_raw_data\\bibek_face.jpg"
    extracted_face_path = detect_and_extract_face(image_path=id_card)
    # face_comparison(extracted_face_path, face_path)