Drowsiness-Detection-ML/demo.py at master · rahdirigs/Drowsiness-Detection-ML · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
import joblib
from utilities import *
from imutils import face_utils
import cv2
import dlib
import numpy as np
import matplotlib.pyplot as plt

knn = joblib.load("model.pkl")
pred = "/Users/rect0r/mlxtend_data/shape_predictor_68_face_landmarks.dat"
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor(pred)
ears = []
mars = []
circs = []
ratios = []
font = cv2.FONT_HERSHEY_SIMPLEX
bottomLeftCornerOfText = (10, 400)
fontScale = 1
fontColor = (255, 255, 255)
lineType = 2


def calibrate():
    data = []
    cap = cv2.VideoCapture(0)
    count = 0
    while True:
        if count == 200:
            break
        _, image = cap.read()
        gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
        rects = detector(image, 0)
        count += 1
        print(count)
        for (i, rect) in enumerate(rects):
            shape = predictor(gray, rect)
            shape = face_utils.shape_to_np(shape)
            data.append(shape)
            cv2.putText(image, "Calibrating...", bottomLeftCornerOfText, font, fontScale, fontColor, lineType)
            for (x, y) in shape:
                cv2.circle(image, (x, y), 2, (0, 255, 0), -1)
        cv2.imshow("Output", image)
        cv2.waitKey(5)

    cv2.destroyAllWindows()
    cap.release()

    base_features = []
    for d in data:
        face = d[36:68]
        ear_value = ear(face)
        mar_value = mar(face)
        circ_value = circularity(face)
        mouth_eye_ratio = mouth_to_eye_ratio(face)
        ears.append(ear_value)
        mars.append(mar_value)
        circs.append(circ_value)
        ratios.append(mouth_eye_ratio)
        base_features.append([ear_value, mar_value, circ_value, mouth_eye_ratio])
    print(base_features)


def get_result(shape):
    face = shape[36:68]
    ear_value = (ear(face) - mean_ear) / std_ear
    mar_value = (mar(face) - mean_ear) / std_mar
    circ = (circularity(face) - mean_circ) / std_circ
    ratio = (mouth_to_eye_ratio(face) - mean_ratio) / std_ratio
    feature = [[ear_value, mar_value, circ, ratio]]
    _result = knn.predict(feature)
    print(_result)
    res_str = "Alert"
    if _result[0] == 10:
        res_str = "Drowsy"
    return res_str, [ear_value, mar_value, circ, ratio]


def live_demo():
    cap = cv2.VideoCapture(0)
    data = []
    result = []
    count = 0
    while True:
        if count == 200:
            break
        _, image = cap.read()
        gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
        rects = detector(image, 0)
        for (i, rect) in enumerate(rects):
            shape = predictor(gray, rect)
            shape = face_utils.shape_to_np(shape)
            result_val, features = get_result(shape)
            count += 1
            print(count)
            data.append(features)
            result.append(result_val)
            cv2.putText(image, result_val, bottomLeftCornerOfText, font, fontScale, fontColor, lineType)
            for (x, y) in shape:
                cv2.circle(image, (x, y), 2, (0, 255, 0), -1)
        cv2.imshow("Output", image)
        cv2.waitKey(300)

    cv2.destroyAllWindows()
    cap.release()
    return data, result


calibrate()
mean_ear = np.mean(ears)
mean_mar = np.mean(mars)
mean_circ = np.mean(circs)
mean_ratio = np.mean(ratios)
std_ear = np.std(ears)
std_mar = np.std(mars)
std_circ = np.std(circs)
std_ratio = np.std(ratios)

print(mean_ear, mean_mar, mean_circ, mean_ratio)
print(std_ear, std_mar, std_circ, std_ratio)

feats, res = live_demo()

plt.title("Results")
plt.xlabel("Frames")
plt.ylabel("State")
plt.plot(res, color="orange")
plt.show()