-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathsplitter.py
196 lines (173 loc) · 7.39 KB
/
splitter.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
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
#!/usr/bin/python
# -*- coding: UTF-8 -*-
# 文字切割 参考:http://chongdata.com/articles/?p=32
import os
import cv2
import numpy as np
import matplotlib.pyplot as plt
import label_dict
import utils
class Splitter(object):
def __init__(self):
print('Create Splitter')
# 计算投影分割点
def extract_peek_ranges_from_array(self, array_vals, minimun_val=10, minimun_range=2):
start_i = None
end_i = None
peek_ranges = []
for i, val in enumerate(array_vals):
if val > minimun_val and start_i is None:
start_i = i
elif val > minimun_val and start_i is not None:
pass
elif val < minimun_val and start_i is not None:
end_i = i
if end_i - start_i >= minimun_range:
peek_ranges.append((start_i, end_i))
start_i = None
end_i = None
elif val < minimun_val and start_i is None:
pass
else:
raise ValueError("cannot parse this case...")
if start_i is not None:
peek_ranges.append((start_i, i + 1))
return peek_ranges
def median_split_ranges(self, peek_ranges):
new_peek_ranges = []
widthes = []
for peek_range in peek_ranges:
w = peek_range[1] - peek_range[0] + 1
widthes.append(w)
widthes = np.asarray(widthes)
median_w = np.median(widthes)
for i, peek_range in enumerate(peek_ranges):
num_char = int(round(widthes[i] / median_w, 0))
if num_char > 1:
char_w = float(widthes[i] / num_char)
for i in range(num_char):
start_point = peek_range[0] + int(i * char_w)
end_point = peek_range[0] + int((i + 1) * char_w)
new_peek_ranges.append((start_point, end_point))
else:
new_peek_ranges.append(peek_range)
return new_peek_ranges
def fill(self, img, i, result_img_path, sub_segment=False):
flag = False
left = 0
right = 0
top = 0
bottom = 0
expect = 30
if img.shape[0] < expect and img.shape[0] > 10:
bottom = int((expect - img.shape[0]) / 2)
top = expect - bottom - img.shape[0]
flag = True
if img.shape[1] < expect and img.shape[1] > 10:
right = int((expect - img.shape[1]) / 2)
left = expect - right - img.shape[1]
flag = True
# if flag:
# img = cv2.copyMakeBorder(img, top, bottom, left, right, cv2.BORDER_CONSTANT, value=[0, 0, 0])
if img.shape[1] < 8 and not sub_segment:
# 数字字符，切割顶部和底部的空隙
# 水平投影
horizontal_sum = np.sum(img, axis=1)
peek_ranges = self.extract_peek_ranges_from_array(horizontal_sum)
img = img[peek_ranges[0][0]:peek_ranges[len(peek_ranges) - 1][1], 0:img.shape[1]]
self.show_img('sub img', img)
# resize resize后效果更好
# img = cv2.resize(img, (30, 30), interpolation=cv2.INTER_CUBIC)
cv2.imwrite(result_img_path + str(i) + '.png', img)
# cv2.imwrite(result_img_path + str(i) + '.png', img)
def process_by_img(self, image_color, result_img_path, minimun_range=11, sub_segment=False, pred_val_list=[]):
# new_shape = (image_color.shape[1] * 2, image_color.shape[0] * 2)
# image_color = cv2.resize(image_color, new_shape)
if sub_segment and len(image_color.shape) == 3:
image_color = cv2.cvtColor(image_color, cv2.COLOR_BGR2GRAY)
if len(image_color.shape) == 2:
adaptive_threshold = image_color
else:
# 黑底白字转换白底黑字
image = cv2.cvtColor(image_color, cv2.COLOR_BGR2GRAY)
adaptive_threshold = cv2.adaptiveThreshold(
image,
255,
cv2.ADAPTIVE_THRESH_GAUSSIAN_C, \
cv2.THRESH_BINARY_INV, 11, 2)
self.show_img('binary image', adaptive_threshold)
# 水平投影
horizontal_sum = np.sum(adaptive_threshold, axis=1)
# plt.plot(horizontal_sum, range(horizontal_sum.shape[0]))
# plt.gca().invert_yaxis()
# plt.show()
peek_ranges = self.extract_peek_ranges_from_array(horizontal_sum)
line_seg_adaptive_threshold = np.copy(adaptive_threshold)
for i, peek_range in enumerate(peek_ranges):
x = 0
y = peek_range[0]
w = line_seg_adaptive_threshold.shape[1]
h = peek_range[1] - y
pt1 = (x, y)
pt2 = (x + w, y + h)
cv2.rectangle(line_seg_adaptive_threshold, pt1, pt2, 255)
self.show_img('line image', line_seg_adaptive_threshold)
vertical_peek_ranges2d = []
for peek_range in peek_ranges:
start_y = peek_range[0]
end_y = peek_range[1]
line_img = adaptive_threshold[start_y:end_y, :]
vertical_sum = np.sum(line_img, axis=0)
vertical_peek_ranges = self.extract_peek_ranges_from_array(
vertical_sum,
minimun_val=30,
minimun_range=minimun_range)
vertical_peek_ranges2d.append(vertical_peek_ranges)
# Draw
counter = 'a'
color = (0, 0, 255)
for i, peek_range in enumerate(peek_ranges):
merge = False
j = 0
c = 0
while j < len(vertical_peek_ranges2d[i]):
if len(pred_val_list) != 0 and utils.resegment(pred_val_list[j]) and not merge:
print('Found a cutting error picture')
merge = True
counter = chr(ord(counter) + 1)
c += 1
j += 1
continue
if merge:
# 识别率低，可能是一个汉字被分割了，合并汉字
x = vertical_peek_ranges2d[i][j - c][0]
if vertical_peek_ranges2d[i][j][1] - x <= 10:
# 分为3部分的字体，eg:柳
c += 1
j += 1
continue
counter = chr(ord(counter) - 1)
merge = False
c = 0
else:
x = vertical_peek_ranges2d[i][j][0]
y = peek_range[0]
w = vertical_peek_ranges2d[i][j][1] - x
h = peek_range[1] - y
pt1 = (x, y)
pt2 = (x + w, y + h)
cv2.rectangle(image_color, pt1, pt2, color)
# 切割图片
sub_img = adaptive_threshold[pt1[1]:pt2[1], pt1[0]:pt2[0]]
self.fill(sub_img, counter, result_img_path, sub_segment=sub_segment)
counter = chr(ord(counter) + 1)
j += 1
self.show_img('char image', image_color)
def process_by_path(self, source_img_path, result_img_path, minimun_range=11, sub_segment=False, pred_val_list=[]):
image_color = cv2.imread(source_img_path)
self.process_by_img(image_color, result_img_path, minimun_range, sub_segment=sub_segment,
pred_val_list=pred_val_list)
def show_img(self, img_name, img):
''
# cv2.imshow(img_name, img)
# cv2.waitKey(0)