dlib 눈 인식(Ear 알고리즘을 이용한 졸음인식)

EAR(Eye Aspect Ratio) 알고리즘은 눈의 비율을 이용해 눈 감김을 확인합니다.

def calculate_EAR(eye): # 눈 거리 계산
	A = distance.euclidean(eye[1], eye[5])
	B = distance.euclidean(eye[2], eye[4])
	C = distance.euclidean(eye[0], eye[3])
	ear_aspect_ratio = (A+B)/(2.0*C)
	return ear_aspect_ratio

EAR 계산

for n in range(36,42): # 오른쪽 눈 감지
        x = face_landmarks.part(n).x
        y = face_landmarks.part(n).y
        leftEye.append((x,y))
        next_point = n+1
        if n == 41:
        	next_point = 36
        x2 = face_landmarks.part(next_point).x
        y2 = face_landmarks.part(next_point).y
        cv2.line(frame,(x,y),(x2,y2),(0,255,0),1)

for n in range(42,48): # 왼쪽 눈 감지
        x = face_landmarks.part(n).x
        y = face_landmarks.part(n).y
        rightEye.append((x,y))
        next_point = n+1
        if n == 47:
        	next_point = 42
        x2 = face_landmarks.part(next_point).x
        y2 = face_landmarks.part(next_point).y
        cv2.line(frame,(x,y),(x2,y2),(0,255,0),1)

눈의 랜드마크 좌표들을 얻어내는 부분입니다.

---

최종 코드

import cv2
import dlib
from functools import wraps
from scipy.spatial import distance
import RPi.GPIO as GPIO
import time

def calculate_EAR(eye): # 눈 거리 계산
	A = distance.euclidean(eye[1], eye[5])
	B = distance.euclidean(eye[2], eye[4])
	C = distance.euclidean(eye[0], eye[3])
	ear_aspect_ratio = (A+B)/(2.0*C)
	return ear_aspect_ratio

# 카메라 셋팅
cap = cv2.VideoCapture(0)
cap.set(3, 640)
cap.set(4, 480)

# dlib 인식 모델 정의
hog_face_detector = dlib.get_frontal_face_detector()
dlib_facelandmark = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat")

# gpio 셋팅
lastsave = 0
output_pin = 18
GPIO.setmode(GPIO.BCM)
GPIO.setup(output_pin, GPIO.OUT, initial = GPIO.LOW)

def counter(func):
    @wraps(func)
    def tmp(*args, **kwargs):
        tmp.count += 1
        time.sleep(0.05)
        global lastsave
        if time.time() - lastsave > 5:
            lastsave = time.time()
            tmp.count = 0
        return func(*args, **kwargs)
    tmp.count = 0
    return tmp

@counter
def close():
    cv2.putText(frame,"DROWSY",(20,100), cv2.FONT_HERSHEY_SIMPLEX,3,(0,0,255),4)

def sound():
    global curr_value
    GPIO.output(output_pin, GPIO.HIGH)
    time.sleep(2) 
    GPIO.output(output_pin, GPIO.LOW)

while True:
    _, frame = cap.read()
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

    faces = hog_face_detector(gray)
    for face in faces:

        face_landmarks = dlib_facelandmark(gray, face)
        leftEye = []
        rightEye = []

        for n in range(36,42): # 오른쪽 눈 감지
        	x = face_landmarks.part(n).x
        	y = face_landmarks.part(n).y
        	leftEye.append((x,y))
        	next_point = n+1
        	if n == 41:
        		next_point = 36
        	x2 = face_landmarks.part(next_point).x
        	y2 = face_landmarks.part(next_point).y
        	cv2.line(frame,(x,y),(x2,y2),(0,255,0),1)

        for n in range(42,48): # 왼쪽 눈 감지
        	x = face_landmarks.part(n).x
        	y = face_landmarks.part(n).y
        	rightEye.append((x,y))
        	next_point = n+1
        	if n == 47:
        		next_point = 42
        	x2 = face_landmarks.part(next_point).x
        	y2 = face_landmarks.part(next_point).y
        	cv2.line(frame,(x,y),(x2,y2),(0,255,0),1)

        left_ear = calculate_EAR(leftEye)
        right_ear = calculate_EAR(rightEye)

        EAR = (left_ear+right_ear)/2
        EAR = round(EAR,2)

        if EAR<0.19:
            close()
            print(f'close count : {close.count}')
            if close.count == 15:
                print("Driver is sleeping")
                sound()
        print(EAR)

    cv2.imshow("Are you Sleepy", frame)

    key = cv2.waitKey(30)
    if key == 27:
        break
        
cap.release()
cv2.destroyAllWindows()

저의 경우 임베디드 보드에 buzzer를 연결하여 경고음이 울리게 했습니다. 

---

http://dlib.net/files/

shape_predictor_68_face_landmarks.dat파일은 위의 링크에서 다운로드하여 사용하시면 됩니다.