Canny_Edge_Detector/Canny.py at master · Samasth/Canny_Edge_Detector · 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
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
import cv2
import numpy as np
import math
import sys

img = cv2.imread(sys.argv[1],cv2.IMREAD_GRAYSCALE)

print(img.shape)
h,w = img.shape
print("height=",h)
print('width=',w)

#Method to perform Gaussian smoothing
def gaussian_smoothing(img):

    # (7*7) gaussian filter for smoothing
    filter1 = np.array([[1,1,2,2,2,1,1],
                        [1,2,2,4,2,2,1],
                        [2,2,4,8,4,2,2],
                        [2,4,8,16,8,4,2],
                        [2,2,4,8,4,2,2],
                        [1,2,2,4,2,2,1],
                        [1,1,2,2,2,1,1]])

    calc = np.zeros((h,w))
    slice2 = np.zeros((7,7))

    # Performing cross convolution operation
    for i in range(3,h-3):
        for j in range(3,w-3):
            slice2 = img[i-3:i+4,j-3:j+4]
            calc[i,j] = np.sum(filter1*slice2)/140
    calc=cv2.convertScaleAbs(calc)
    cv2.imshow("gauss", calc)
    cv2.imwrite('gaussian_smoothing.png', calc)
            #cv2.waitKey()
    return calc

# Method to perform Gradiant operations
def gradient (calc):

    calcx = np.zeros((h,w))
    calcy = np.zeros((h,w))
    calcz = np.zeros((h,w))
    calca = np.zeros((h,w))
    calcxo = np.zeros((h,w))
    calcyo = np.zeros((h,w))
    #np.set_printoptions(suppress=True)

    gx = np.array([[-1,0,1],[-2,0,2],[-1,0,1]])                 # sobel operator for horizontal gradiant
    gy = np.array([[1,2,1],[0,0,0],[-1,-2,-1]])                 # sobel operator for vertical gradiant
    slice3 = np.zeros((3,3))
    # Performing Gradiant operations
    for i in range(4,h-4):
        for j in range(4,w-4):
            slice3 = calc[i-1:i+2,j-1:j+2]
            # Computing horizontal and vertical gradiants
            calcxo[i,j] = (np.sum(gx*slice3))
            calcyo[i,j] = (np.sum(gy*slice3))

    x_max = np.amax(calcxo)
    y_max = np.amax(calcyo)

    # Performing normalisation to keep the pixel values in the range 0-255
    np.divide(calcxo, x_max)
    np.divide(calcyo, y_max)
    np.multiply(calcxo, 255)
    np.multiply(calcyo, 255)

    for i in range(4,h-4):
        for j in range(4,w-4):
            # Calculating Gradiant Magnitude
            slice3 = calc[i-1:i+2,j-1:j+2]
            calcx[i,j] = (np.sum(gx*slice3))**2
            calcy[i,j] = (np.sum(gy*slice3))**2
            calcz[i,j] = np.sqrt(calcx[i,j]+calcy[i,j])
            #Calculating Gradiant Angle
            if calcxo[i,j] != 0:
                calca[i,j] = math.degrees(math.atan(calcyo[i,j]/calcxo[i,j]))
            else:
                if calcyo[i,j] != 0:
                    calca[i,j] = 90.0
                else:
                    calca[i,j] = 0
    # Calculating absolute values of horizontal and vertical gradiants
    calcxo = np.absolute(calcxo)
    calcyo = np.absolute(calcyo)

    # Displaying Horizontal and Vertical Gradiants image
    calcxo = cv2.convertScaleAbs(calcxo)
    cv2.imshow("Vertical gradiant", calcxo)
    cv2.imwrite('vertical_gradiant.png', calcxo)
    calcyo = cv2.convertScaleAbs(calcyo)
    cv2.imshow("Horizontal gradiant", calcyo)
    cv2.imwrite('horizontal_gradiant.png', calcyo)

    # Displaying Gradiant magnitude image
    calcz = cv2.convertScaleAbs(calcz)
    cv2.imshow("gradiant Magnitude", calcz)
    cv2.imwrite('gradiant_magnitude.png', calcz)

    return calca,calcz


# Method to perform Non-Maxima suppression
def non_maxima_supression(calca,calcz):

    calcb = np.zeros((h-4,w-4))
    for i in range(4,h-4):
        for j in range(4,w-4):
            # Assigning gradiant angles to sectors in the range 0-3
            if calca[i,j] > -22.5 and calca[i,j] <= 22.5:
                calcb[i,j] = 0
            elif calca[i,j] > 22.5 and calca[i,j] <= 67.5:
                calcb[i,j] = 1
            elif calca[i,j] > 67.5 and calca[i,j] <= 112.5:
                calcb[i,j] = 2
            elif calca[i,j] > -67.5 and calca[i,j] <= -22.5:
                calca[i,j] = 3

    # Checking if the Gradiant magnitude of center value is greater than its adjacent values along the gradiant angle
    calcc = np.zeros((h,w))
    for i in range(5,h-5):
        for j in range(5,w-5):
            if calcb[i,j] == 0:
                if calcz[i,j]>calcz[i,j-1] and calcz[i,j]>calcz[i,j+1]:
                    calcc[i,j] = calcz[i,j]
                else:
                    calcc[i,j] = 0
            elif calcb[i,j] == 1:
                if calcz[i,j]>calcz[i-1,j+1] and calcz[i,j]>calcz[i+1,j-1]:
                    calcc[i,j] = calcz[i,j]
                else:
                    calcc[i,j] = 0
            elif calcb[i,j] == 2:
                if calcz[i,j]>calcz[i-1,j] and calcz[i,j]>calcz[i+1,j]:
                    calcc[i,j] = calcz[i,j]
                else:
                    calcc[i,j] = 0
            elif calcb[i,j] == 3:
                if calcz[i,j]>calcz[i-1,j-1] and calcz[i,j]>calcz[i+1,j+1]:
                    calcc[i,j] = calcz[i,j]
                else:
                    calcc[i,j] = 0

    calcc = cv2.convertScaleAbs(calcc)
    cv2.imshow("After non-maxima suppression", calcc)
    cv2.imwrite('non_maxima_supp.png', calcc)
    return calcc,calca

# Method to perform double thresholding
def double_thresholding (calcc,calca):
    # selecting two threshold values
    t1 = 60
    t2 = 100
    # dividing the image into three parts based on threshold values t1 and t2
    for i in range(5,h-5):
        for j in range(5,w-5):
            if calcc[i,j] < t1:
                calcc[i,j] = 0
            elif calcc[i,j] > t2:
                calcc[i,j] = 255
            else:
                if max(calcc[i-1,j-1],calcc[i-1,j],calcc[i-1,j+1],calcc[i,j-1],calcc[i,j+1],calcc[i+1,j-1],calcc[i+1,j],calcc[i+1,j+1])>t2 or min(abs(calca[i,j]-calca[i-1,j-1]),abs(calca[i,j]-calca[i-1,j]),abs(calca[i,j]-calca[i-1,j+1]),abs(calca[i,j]-calca[i,j-1]),abs(calca[i,j]-calca[i,j+1]),abs(calca[i,j]-calca[i+1,j-1]),abs(calca[i,j]-calca[i+1,j]),abs(calca[i,j]-calca[i+1,j+1])) <=45:
                    calcc[i,j] = 255

    # second pass
    for i in range(5,h-5):
        for j in range(5,w-5):
            if calcc[i,j] <= t2 and calcc[i,j] >= t1:
                calcc[i,j] = 0

    cv2.imwrite('thresholding.png', calcc)
    cv2.imshow("Final image after thresholding", calcc)
    cv2.waitKey()


calcf = gaussian_smoothing(img)
calcaf,calczf = gradient(calcf)
calccf,calcaf = non_maxima_supression(calcaf,calczf)
double_thresholding(calccf,calcaf)