-
Notifications
You must be signed in to change notification settings - Fork 4
/
Split_Words.py
162 lines (117 loc) · 5.8 KB
/
Split_Words.py
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
import cv2
import numpy as np
def Sorting_Key(rect):
global Lines, Size
x, y, w, h = rect
cx = x + int(w / 2)
cy = y + int(h / 2)
for i, (upper, lower) in enumerate(Lines):
if not any([all([upper > y + h, lower > y + h]), all([upper < y, lower < y])]):
return cx + ((i + 1) * Size)
def Split(Image):
global Lines, Size
gray = cv2.cvtColor(Image, cv2.COLOR_BGR2GRAY)
kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (25, 25))
morph = cv2.morphologyEx(gray, cv2.MORPH_CLOSE, kernel)
for i in range(morph.shape[0]):
for j in range(morph.shape[1]):
if not morph[i][j]:
morph[i][j] = 1
div = gray / morph
gray = np.array(cv2.normalize(div, div, 0, 255, cv2.NORM_MINMAX), np.uint8)
_, thresh = cv2.threshold(gray, 180, 255, cv2.THRESH_BINARY_INV)
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (3, 3))
thresh = cv2.morphologyEx(thresh, cv2.MORPH_DILATE, kernel, iterations = 1)
contours, _ = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)
i = 0
Length = len(contours)
while i < Length:
x, y, w, h = cv2.boundingRect(contours[i])
if w * h <= 100:
del contours[i]
i -= 1
Length -= 1
i += 1
h_proj = np.sum(thresh, axis = 1)
upper = None
lower = None
Lines = []
for i in range(h_proj.shape[0]):
proj = h_proj[i]
if proj != 0 and upper == None:
upper = i
elif proj == 0 and upper != None and lower == None:
lower = i
if lower - upper >= 30:
Lines.append([upper, lower])
upper = None
lower = None
if upper:
Lines.append([upper, h_proj.shape[0] - 1])
Size = thresh.shape[1]
bounding_rects = []
for contour in contours:
x, y, w, h = cv2.boundingRect(contour)
for upper, lower in Lines:
if not any([all([upper > y + h, lower > y + h]), all([upper < y, lower < y])]):
bounding_rects.append([x, y, w, h])
i = 0
Length = len(bounding_rects)
while i < Length:
x, y, w, h = bounding_rects[i]
j = 0
while j < Length:
distancex = abs(bounding_rects[j][0] - (bounding_rects[i][0] + bounding_rects[i][2]))
distancey = abs(bounding_rects[j][1] - (bounding_rects[i][1] + bounding_rects[i][3]))
threshx = max(abs(bounding_rects[j][0] - (bounding_rects[i][0] + bounding_rects[i][2])),
abs(bounding_rects[j][0] - bounding_rects[i][0]),
abs((bounding_rects[j][0] + bounding_rects[j][2]) - bounding_rects[i][0]),
abs((bounding_rects[j][0] + bounding_rects[j][2]) - (bounding_rects[i][0] + bounding_rects[i][2])))
threshy = max(abs(bounding_rects[j][1] - (bounding_rects[i][1] + bounding_rects[i][3])),
abs(bounding_rects[j][1] - bounding_rects[i][1]),
abs((bounding_rects[j][1] + bounding_rects[j][3]) - bounding_rects[i][1]),
abs((bounding_rects[j][1] + bounding_rects[j][3]) - (bounding_rects[i][1] + bounding_rects[i][3])))
if i != j and any([all([not any([all([bounding_rects[j][1] > y + h, bounding_rects[j][1] + bounding_rects[j][3] > y + h]), all([bounding_rects[j][1] < y, bounding_rects[j][1] + bounding_rects[j][3] < y])]),
not any([all([bounding_rects[j][0] > x + w, bounding_rects[j][0] + bounding_rects[j][2] > x + w]), all([bounding_rects[j][0] < x, bounding_rects[j][0] + bounding_rects[j][2] < x])])]),
all([distancex <= 10, bounding_rects[i][3] + bounding_rects[j][3] + 10 >= threshy]), all([bounding_rects[i][2] + bounding_rects[j][2] + 10 >= threshx, distancey <= 10])]):
x = min(bounding_rects[i][0], bounding_rects[j][0])
w = max(bounding_rects[i][0] + bounding_rects[i][2], bounding_rects[j][0] + bounding_rects[j][2]) - x
y = min(bounding_rects[i][1], bounding_rects[j][1])
h = max(bounding_rects[i][1] + bounding_rects[i][3], bounding_rects[j][1] + bounding_rects[j][3]) - y
bounding_rects[i] = [x, y, w, h]
del bounding_rects[j]
i = -1
Length -= 1
break
j += 1
i += 1
bounding_rects.sort(key = Sorting_Key)
Words = []
for x, y, w, h in bounding_rects:
crop = Image[y:y + h, x:x+ w]
gray = cv2.cvtColor(crop, cv2.COLOR_BGR2GRAY)
kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (25, 25))
morph = cv2.morphologyEx(gray, cv2.MORPH_CLOSE, kernel)
for i in range(morph.shape[0]):
for j in range(morph.shape[1]):
if not morph[i][j]:
morph[i][j] = 1
div = gray / morph
gray = np.array(cv2.normalize(div, div, 0, 255, cv2.NORM_MINMAX), np.uint8)
_, thresh = cv2.threshold(gray, 180, 255, cv2.THRESH_BINARY_INV)
contours, _ = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)
contours = np.vstack(contours)
rect = cv2.minAreaRect(contours)
Box = cv2.boxPoints(rect)
index = np.argmin(np.sum(Box, axis = 1))
box = []
box.extend(Box[index:])
box.extend(Box[0:index])
box = np.int0(box)
shape = (box[1][0] - box[0][0], box[3][1] - box[0][1])
src = np.float32(box)
dst = np.array([[0, 0], [shape[0], 0], [shape[0], shape[1]], [0, shape[1]]], np.float32)
M = cv2.getPerspectiveTransform(src, dst)
warp = cv2.bitwise_not(cv2.warpPerspective(cv2.bitwise_not(crop), M, shape))
Words.append(warp.copy())
return Words