/
ocr.py
executable file
·238 lines (200 loc) · 8.06 KB
/
ocr.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
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
#!/usr/bin/python
#
# Copyright (c) 2012 Ludwig Haehne
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.
from argparse import ArgumentParser
from glob import glob
import logging
import os
from Queue import Queue
from shutil import rmtree
from subprocess import check_call, check_output, Popen, PIPE
from tempfile import mkdtemp
from threading import Thread
import time
logging.basicConfig(level=logging.DEBUG,
format='%(asctime)s [%(levelname)s] %(message)s',
datefmt='%H:%M:%S',
)
class Timer:
def __enter__(self):
self.start = time.time()
return self
def __exit__(self, *args):
self.end = time.time()
self.interval = self.end - self.start
def system_info():
deps = []
pdfimages = Popen(['pdfimages', '-v'], stderr=PIPE).communicate()[1]
deps.append(('pdfimages', pdfimages.split('\n')[0]))
tesseract = Popen(['tesseract', '-v'], stderr=PIPE).communicate()[1]
deps.append(('tesseract', tesseract.split('\n')[0]))
# hocr2pdf doesn't support a --version flag but prints the version to
# stderr if called without arguments
hocr2pdf = Popen(['hocr2pdf', '--help'], stdout=PIPE, stderr=PIPE)
hocr2pdf = hocr2pdf.communicate()[1]
hocr2pdf = [line for line in hocr2pdf if 'version' in line]
if hocr2pdf:
deps.append(('hocr2pdf', hocr2pdf))
gs = check_output(['gs', '--version']).split('\n')[0]
deps.append(('gs', gs))
identify = check_output(['identify', '--version']).split('\n')[0]
deps.append(('identify', identify))
convert = check_output(['convert', '--version']).split('\n')[0]
deps.append(('convert', convert))
return deps
def extract_images(pdf, output):
check_call(["pdfimages", pdf, output])
images = []
for filetype in ('*.ppm', '*.pbm', '*.jpg'):
images.extend(glob(os.path.join(output, filetype)))
images.sort()
return images
def compute_dpi(pdf_w, pdf_h, image_w, image_h):
"""
Deduce scan resolution from PDF and image size.
http://stackoverflow.com/a/576816/63392
"""
dpi_w = int(round(image_w*72./pdf_w))
dpi_h = int(round(image_h*72./pdf_h))
return dpi_w, dpi_h
def get_resolution(filename):
"""
Return resolution per page.
"""
pages = check_output(["identify", "-format", "%w,%h;", filename])
pages = [page.split(',') for page in pages.split(';') if page.strip()]
pages = [(int(x), int(y)) for x, y in pages]
return pages
PAPER_SIZES = {
(841, 1189): 'A0',
(594, 841): 'A1',
(420, 594): 'A2',
(297, 420): 'A3',
(210, 297): 'A4',
(148, 210): 'A5',
(105, 148): 'A6',
( 74, 105): 'A7',
( 52, 74): 'A8',
(250, 353): 'B4',
(176, 250): 'B5',
(125, 176): 'B6',
(215, 279): 'US Letter',
}
def get_paper_type(wdots, hdots):
w_mm = int(wdots / 7.2 * 2.54)
h_mm = int(hdots / 7.2 * 2.54)
if (w_mm, h_mm) in PAPER_SIZES:
return PAPER_SIZES[(w_mm, h_mm)]
elif (h_mm, w_mm) in PAPER_SIZES:
return PAPER_SIZES[(h_mm, w_mm)] + ', landscape'
else:
return ''
def ocr_page(image, lang='eng', width=-1, height=-1):
base = os.path.splitext(image)[0]
png = base + '.png'
hocr = base + '.html'
pdf = base + '.pdf'
w, h = get_resolution(image)[0]
dpi_w, dpi_h = compute_dpi(width, height, w, h)
logging.debug("Page={}x{} Image={}x{} DPI={}x{}".format(
width, height, w, h, dpi_w, dpi_h))
check_call(["convert", image, png])
devnull = open('/dev/null', 'w')
check_call(["tesseract", png, base, '-l', lang, 'hocr'],
stdout=devnull)
# Reduce resolution of images to 1/2 that have 600dpi or more.
if dpi_w >= 600:
check_call(["mogrify", "-resize", "50%", png])
dpi_w = dpi_w / 2
html = os.open(hocr, os.O_RDONLY)
check_call(['hocr2pdf', '-r', str(dpi_w), '-i', png, '-o', pdf],
stdin=html)
os.close(html)
return pdf
def process_page(index, queue, lang, resolution):
"""
Pull page from the queue and perform OCR. Make sure to acknowledge the
message in the queue even if there is an exception so that the main thread
does not block on a queue that will never be empty.
"""
while True:
page, image = queue.get()
try:
logging.info("Page {:>2}: Run OCR ...".format(page))
width, height = resolution[page-1]
with Timer() as t:
ocr_page(image, lang=lang, width=width, height=height)
logging.info("Page {:>2}: OCR took {:.2f}s".format(page, t.interval))
finally:
queue.task_done()
def merge_pdf(pages, output_filename):
check_call(['gs',
'-q',
'-dNOPAUSE',
'-dBATCH',
'-sDEVICE=pdfwrite',
'-dCompatibilityLevel=1.4',
'-sOutputFile={}'.format(output_filename)] +
pages,
)
def start_workers(num_workers, queue, lang, resolution):
for tid in range(num_workers):
args = (tid, queue, lang, resolution)
worker = Thread(target=process_page, args=args)
worker.daemon = True
worker.start()
def process(input_file, output_file, lang='eng', jobs=4):
tmp = os.path.join(mkdtemp(), '')
try:
resolution = get_resolution(input_file)
w, h = resolution[0]
logging.info("{} pages, {}mm*{}mm {}".format(len(resolution),
int(w / 7.2 * 2.54),
int(h / 7.2 * 2.54),
get_paper_type(w, h)))
logging.info("Extract pages from {}".format(input_file))
images = extract_images(input_file, tmp)
num_workers = min(len(images), jobs)
queue = Queue()
start_workers(num_workers, queue, lang, resolution)
logging.info("Process {} pages with {} threads".format(len(images),
num_workers))
for idx, image in enumerate(images, start=1):
queue.put((idx, image))
queue.join()
pages = sorted(glob(os.path.join(tmp, '*.pdf')))
logging.info("OCR complete. Merge pages into '{}'".format(output_file))
merge_pdf(pages, output_file)
check_call(['ls', '-lh', input_file, output_file])
finally:
rmtree(tmp)
if __name__ == '__main__':
parser = ArgumentParser(description="Add OCR to overlay to scanned PDF")
parser.add_argument('input', nargs=1, help='Scanned PDF')
parser.add_argument('output', nargs=1, help='Output PDF')
parser.add_argument('-l', '--lang', default='eng',
help='3-digit tesseract language code (default "eng")')
parser.add_argument('-j', '--jobs', default=4, type=int,
help='Specifies the number of pages to process simultaneously')
args = parser.parse_args()
for name, version in system_info():
logging.info('{:<12}: {}'.format(name, version))
process(args.input[0], args.output[0], lang=args.lang, jobs=args.jobs)