Iris-positioning/main.py at master · f31n/Iris-positioning · 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
129
130
131
132
133
134
135
136
# Localização da iris por Hough Cirles e
# indicação da direção de visão.
#
# Trabalho T5
#
# Aluno: Eliel Marcos Romancini - 15205574
#

import cv2
import numpy as np
import math


def nothing(x):
    pass


sizex = 500
sizey = 300

cam = cv2.VideoCapture('Eyes.mp4')
out = cv2.VideoWriter('output.mp4', -1, 30.0, (500, 300))

ret, frame = cam.read()

while ret:

    frame = cv2.resize(frame, (sizex, sizey))

    frameT = frame.copy()

    cv2.putText(frame, 'Iris Positioning - Eliel Romancini', (250, 295), cv2.FONT_HERSHEY_PLAIN, 0.9, (0, 0, 0))

    frame_pupil = cv2.medianBlur(frame, 19)

    frame_pupil_gray = cv2.cvtColor(frame_pupil, cv2.COLOR_BGR2GRAY)

    frame_eye = cv2.GaussianBlur(frame, (55, 55), sigmaX=0, sigmaY=0)

    frame_eye_hsv = cv2.cvtColor(frame_eye, cv2.COLOR_BGR2HSV)

    frame_eye_bin1 = cv2.inRange(frame_eye_hsv, (0, 45, 4), (15, 255, 255))
    frame_eye_bin2 = cv2.inRange(frame_eye_hsv, (175, 45, 4), (179, 255, 255))
    frame_eye_bin = frame_eye_bin1 + frame_eye_bin2
    frame_eye_bin = cv2.morphologyEx(frame_eye_bin, cv2.MORPH_OPEN, np.ones((3, 3), np.uint8))
    frame_eye_bin = cv2.morphologyEx(frame_eye_bin, cv2.MORPH_DILATE, np.ones((3, 3), np.uint8))

    frame_eye_bin = cv2.bitwise_not(frame_eye_bin)

    #  Encontrar o contorno do olho
    contorns = cv2.findContours(frame_eye_bin, mode=cv2.RETR_TREE, method=cv2.CHAIN_APPROX_NONE)[0]
    circles = []
    center = 25
    offset = 25

    if not (contorns == []):
        cnt = max(contorns, key=cv2.contourArea)
        cnt_approx = cv2.convexHull(cnt)
        cnt_moments = cv2.moments(cnt_approx)
        cnt_area = cnt_moments['m00']
        if cnt_area > 12000:
            #  Encontrou olho
            EyeX = int(cnt_moments['m10'] / cnt_area)
            EyeY = int(cnt_moments['m01'] / cnt_area)
            cv2.drawContours(frame, [cnt_approx], -1, (0, 255, 255), 2, cv2.LINE_AA)
            cv2.circle(frame, (EyeX, EyeY), center, (255, 0, 0), 1, cv2.LINE_AA)
            cv2.line(frame, (EyeX, 0 + offset), (EyeX, sizey - offset), (255, 0, 0), 1, cv2.LINE_AA)
            cv2.line(frame, (0 + offset, EyeY), (sizex - offset, EyeY), (255, 0, 0), 1, cv2.LINE_AA)

            #  Encontrar Pupila
            circles = cv2.HoughCircles(frame_pupil_gray, cv2.HOUGH_GRADIENT, dp=1, minDist=1000, param1=90, param2=20,
                                       minRadius=15, maxRadius=38)

            if (circles is None) or (circles == []):
                cv2.imshow('video', frame)
                out.write(frame)
                key = cv2.waitKey(1)
                if key == 27:
                    break

                ret, frame = cam.read()
                continue

            #  Encontrou Pupila
            for (PupilX, PupilY, PupilR) in circles[0, :]:
                PupilX = int(PupilX)
                PupilY = int(PupilY)
                PupilR = int(PupilR)
                cv2.circle(frame, (PupilX, PupilY), PupilR, (0, 255, 0), 2, cv2.LINE_AA)
                cv2.line(frame, (PupilX, PupilY), (EyeX, EyeY), (255, 0, 255), 2, cv2.LINE_AA)

                # Calcula angulo em relação ao eixo X
                vectorX = [1, 0]
                vectorPE = [PupilX - EyeX, PupilY - EyeY]
                if vectorPE == [0, 0]:
                    angle = 0
                else:
                    angle = math.acos(((vectorX[0] * vectorPE[0]) + (vectorX[1] * vectorPE[1])) / (
                            math.hypot(vectorX[0], vectorX[1]) * math.hypot(vectorPE[0], vectorPE[1]))) * 180 / math.pi
                if vectorPE[1] > 0:
                    angle = 360 - angle

                #  Calcula distância entre centro do olho e centro da pupila
                dist = ((PupilX - EyeX) ** 2 + (PupilY - EyeY) ** 2) ** (1 / 2)

                if dist < center:
                    cv2.putText(frame, 'CENTER', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))
                else:
                    if angle > 15 and angle <= 75:
                        cv2.putText(frame, 'TOP / LEFT', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))
                    if angle > 75 and angle <= 105:
                        cv2.putText(frame, 'TOP', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))
                    if angle > 105 and angle <= 165:
                        cv2.putText(frame, 'TOP / RIGHT', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))
                    if angle > 165 and angle <= 195:
                        cv2.putText(frame, 'RIGHT', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))
                    if angle > 195 and angle <= 255:
                        cv2.putText(frame, 'DOWN / RIGHT', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))
                    if angle > 255 and angle <= 285:
                        cv2.putText(frame, 'DOWN', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))
                    if angle > 285 and angle <= 345:
                        cv2.putText(frame, 'DOWN / LEFT', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))
                    if angle > 345 or angle <= 15:
                        cv2.putText(frame, 'LEFT', (10, 30), cv2.FONT_HERSHEY_PLAIN, 1.8, (255, 255, 255))

    cv2.imshow('video', frame)
    out.write(frame)
    key = cv2.waitKey(1)
    if key == 27:
        break

    ret, frame = cam.read()

cam.release()
out.release()
cv2.destroyAllWindows()