-
Notifications
You must be signed in to change notification settings - Fork 0
/
train.py
145 lines (116 loc) · 5.24 KB
/
train.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
import sys
import operator
import numpy as np
import cv2
def check_for_font(font_h):
max,font_size = -1, 0
for i in font_h:
count=0
for j in font_h:
if abs(j-i)<10:
count+=1
if count>max:
max=count
font_size=i
return font_size
def ocr_start(contours,im,count):
samples = np.empty((0,1600))
responses = []
keys = [i for i in range(48,58)]
# samples = np.loadtxt('train/generalsamples.data',np.float32)
# responses = np.loadtxt('train/generalresponses.data',np.float32)
# responses = responses.reshape((responses.size,1))
out = np.zeros(im.shape,np.uint8)
height=im.shape[0]
font_h=[]
thresh = height/9
contour_rect=[]
for cnt in contours:
if cv2.contourArea(cnt)>50:
[x,y,w,h] = cv2.boundingRect(cnt)
if h<thresh-10 and h>thresh/4 and w>thresh/6 and w<thresh-10 :
font_h.append(h)
contour_rect.append([x,y,w,h])
font_size = check_for_font(font_h)
for rect in contour_rect:
[x,y,w,h] = rect
if abs(h-font_size) < 15:
cv2.rectangle(im,(x,y),(x+w,y+h),(255,0,0),1)
roi = im[y:y+h,x:x+w]
roismall = cv2.resize(roi,(40,40))
cv2.imshow('im',im)
cv2.imshow('norm',roismall)
key = cv2.waitKey(0)
if key == 27: # (escape to quit)
sys.exit()
elif key in keys:
responses.append(int(chr(key)))
print(roismall)
sample = roismall.reshape((1,1600))
print(sample)
samples = np.append(samples,sample,0)
print key
responses = np.array(responses,np.float32)
responses = responses.reshape((responses.size,1))
print "training complete"
np.savetxt('generalsamples'+str(count)+'.data',samples)
np.savetxt('generalresponses'+str(count)+'.data',responses)
def distance(a, b):
return np.sqrt( ((a[0] - b[0]) **2) + ((a[1] - b[1]) **2) )
def extract_contour(image):
contours, h = cv2.findContours(image, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
return contours
def CutoutMaxContour(image):
contours = extract_contour(image)
return getContourCorners(max(contours, key=cv2.contourArea))
def show(img, windowName='Image'):
screen_res = 1280.0, 720.0
scale_width = screen_res[0] / img.shape[1]
scale_height = screen_res[1] / img.shape[0]
scale = min(scale_width, scale_height)
window_width = int(img.shape[1] * scale)
window_height = int(img.shape[0] * scale)
cv2.namedWindow(windowName, cv2.WINDOW_NORMAL)
cv2.resizeWindow(windowName, window_width, window_height)
cv2.imshow(windowName, img)
cv2.waitKey(0)
cv2.destroyAllWindows()
def getContourCorners(contour):
bottom_right, _ = max(enumerate([point[0][0] + point[0][1] for point in contour]), key=operator.itemgetter(1))
bottom_left , _ = max(enumerate([point[0][0] - point[0][1] for point in contour]), key=operator.itemgetter(1))
top_right , _ = min(enumerate([point[0][0] - point[0][1] for point in contour]), key=operator.itemgetter(1))
top_left , _ = min(enumerate([point[0][0] + point[0][1] for point in contour]), key=operator.itemgetter(1))
sudoku_corners=[contour[top_left][0],contour[top_right][0],contour[bottom_left][0],contour[bottom_right][0]]
return sudoku_corners
def Linear_transform_image(image, crop_rectangle):
side = max(distance(pointA,pointB) for pointA in crop_rectangle for pointB in crop_rectangle)
source_polygon = np.array(crop_rectangle, dtype='float32')
dest_square = np.array([[0, 0],[0, side - 1],[side - 1, 0],[side - 1, side - 1],], dtype='float32')
m = cv2.getPerspectiveTransform(source_polygon, dest_square)
return cv2.warpPerspective(image, m, (int(side), int(side)))
def preprocess(image):
image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (2, 2))
image = cv2.morphologyEx(image, cv2.MORPH_CLOSE, kernel)
blur=len(image)/100
blur = blur+(blur+1)%2;
image = cv2.GaussianBlur(image,(blur,blur), 0)
image = 255 - cv2.adaptiveThreshold(image, 255, cv2.ADAPTIVE_THRESH_MEAN_C,
cv2.THRESH_BINARY, 17, 5)
# show(image)
return image
def main():
count=0
for i in range (1,15):
count+=1
img = cv2.imread('sample/image'+str(i)+'.jpg')
# show(img)
preprocessed_image = preprocess(img)
sudoku_corners = CutoutMaxContour(preprocessed_image)
sudoku_img = Linear_transform_image(img, sudoku_corners)
# show(sudoku_img)
# cv2.imwrite('train/train.jpg',sudoku_img)
sudoku_img = preprocess(sudoku_img)
ocr_start(extract_contour(sudoku_img) ,sudoku_img,count)
if __name__ == '__main__':
main()