face_detection.py

import cv2
import numpy as np

# Resize back to original size
def resize_to_original(frame, width, height):
    return cv2.resize(frame, (width, height), interpolation=cv2.INTER_AREA)

def face_detection(frame, detector, use_larger_box=False, larger_box_coef=1.0):
    """Face detection on a single frame.

    Args:
        frame(np.array): a single frame.
        use_larger_box(bool): whether to use a larger bounding box on face detection.
        larger_box_coef(float): Coef. of larger box.
    Returns:
        face_box_coor(List[int]): coordinates of face bouding box.
    """
    detector = cv2.CascadeClassifier('./utils/haarcascade_frontalface_default.xml')
    face_zone = detector.detectMultiScale(frame)
    if len(face_zone) < 1:
        print("ERROR: No Face Detected")
        face_box_coor = [0, 0, frame.shape[0], frame.shape[1]]
    elif len(face_zone) >= 2:
        face_box_coor = np.argmax(face_zone, axis=0)
        face_box_coor = face_zone[face_box_coor[2]]
        print("Warning: More than one faces are detected(Only cropping the biggest one.)")
    else:
        face_box_coor = face_zone[0]
    if use_larger_box:
        face_box_coor[0] = max(0, face_box_coor[0] - (larger_box_coef - 1.0) / 2 * face_box_coor[2])
        face_box_coor[1] = max(0, face_box_coor[1] - (larger_box_coef - 1.0) / 2 * face_box_coor[3])
        face_box_coor[2] = larger_box_coef * face_box_coor[2]
        face_box_coor[3] = larger_box_coef * face_box_coor[3]
    return face_box_coor