Spaces:
Sleeping
Sleeping
| import cv2 | |
| import base64 | |
| from utils import * | |
| CIN = ["Nom complet", "Date de naissance", "Date de validité", "Lieu de naissance", "Numéro CIN"] | |
| def extract_face(image, scale_factor=1.2): | |
| # Load the pre-trained face detector | |
| face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_alt.xml') | |
| # Convert the image to grayscale | |
| gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) | |
| # Detect faces in the image | |
| faces = face_cascade.detectMultiScale(gray, scaleFactor=scale_factor, minNeighbors=5, minSize=(30, 30)) | |
| print(len(faces)) | |
| if len(faces) == 0: | |
| print("No faces found") | |
| return None | |
| # Assume only one face in the image for simplicity | |
| (x, y, w, h) = max(faces, key=lambda x: x[2]) | |
| # Expand the bounding box to include a slightly larger region | |
| x -= int(0.1 * w) | |
| y -= int(0.1 * h) | |
| w += int(0.2 * w) | |
| h += int(0.2 * h) | |
| # Ensure the coordinates are within the image boundaries | |
| x = max(x, 0) | |
| y = max(y, 0) | |
| w = min(w, image.shape[1]) | |
| h = min(h, image.shape[0]) | |
| # Extract the face region from the image | |
| face = image[y:y+h, x:x+w] | |
| # Convert the face to base64 string | |
| _, encoded_image = cv2.imencode('.jpg', face) | |
| encoded_image_str = base64.b64encode(encoded_image).decode('utf-8') | |
| # Create a dictionary to store image information | |
| image_info = { | |
| "width": face.shape[1], | |
| "height": face.shape[0], | |
| "data": encoded_image_str | |
| } | |
| return image_info | |
| if __name__ == '__main__': | |
| lang = "french" | |
| to_be_extracted = CIN | |
| image_path = "./docs for ocr/CIN 2.png" | |
| image = cv2.imread(image_path) | |
| text_data = extract_data(lang, to_be_extracted, image) | |
| print(text_data) | |
| face_data = extract_face(image) |