forked from ocropus/hocr-tools
-
Notifications
You must be signed in to change notification settings - Fork 0
/
hocr-extract-images
executable file
·115 lines (102 loc) · 3.11 KB
/
hocr-extract-images
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
#!/usr/bin/env python
# extract the images and texts within all the ocr_line elements
# within the hOCR file
from __future__ import print_function
import argparse
import codecs
import os
import re
import sys
from lxml import html
from PIL import Image
def get_text(node):
textnodes = node.xpath('.//text()')
s = ''.join([text for text in textnodes])
return re.sub(r'\s+', ' ', s)
def get_prop(node, name):
title = node.get("title")
props = title.split(';')
for prop in props:
(key, args) = prop.split(None, 1)
args = args.strip('"')
if key == name:
return args
return None
parser = argparse.ArgumentParser(
description=("Extract the images and texts within all the ocr_line "
"elements within the hOCR file")
)
parser.add_argument(
"file",
help="hOCR file",
type=argparse.FileType('r'),
nargs='?',
default=sys.stdin)
parser.add_argument("-b", "--basename", help="image-dir")
parser.add_argument(
"-p",
"--pattern",
help="file-pattern, default: %(default)s",
default="line-%03d.png")
parser.add_argument(
"-e",
"--element",
help="element-name, default: %(default)s",
default="ocr_line")
parser.add_argument(
"-P",
"--pad",
default=None,
help="extra padding for bounding box")
parser.add_argument(
"-U",
"--unicodedammit",
action="store_true",
help="attempt to use BeautifulSoup.UnicodeDammit to fix encoding issues")
args = parser.parse_args()
padding = None
if args.pad is not None:
padding = eval("["+args.pad+"]")
assert len(padding) in [1, 4], (args.pad, padding)
if len(padding) == 1:
padding = padding * 4
tpattern = args.pattern + '.txt'
if args.pattern[-4] == '.':
tpattern = args.pattern[:-3] + 'txt'
if args.unicodedammit:
from bs4 import UnicodeDammit
content = args.file.read().encode()
doc = UnicodeDammit(content, is_html=True)
parser = html.HTMLParser(encoding=doc.original_encoding)
doc = html.document_fromstring(content, parser=parser)
else:
doc = html.parse(args.file)
pages = doc.xpath('//*[@class="ocr_page"]')
for page in pages:
iname = get_prop(page, 'file')
if not iname:
iname = get_prop(page, 'image')
if args.basename:
iname = os.path.join(args.basename, os.path.basename(iname))
if not os.path.exists(iname):
print("not found:", iname)
sys.exit(1)
image = Image.open(iname)
lines = page.xpath("//*[@class='%s']" % args.element)
lcount = 1
for line in lines:
bbox = [int(x) for x in get_prop(line, 'bbox').split()]
if padding is not None:
w, h = image.size
bbox[0] = max(bbox[0]-padding[0], 0)
bbox[1] = max(bbox[1]-padding[1], 0)
bbox[2] = min(bbox[2]+padding[2], w)
bbox[3] = min(bbox[3]+padding[3], h)
if bbox[0] > bbox[2] or bbox[1] >= bbox[3]:
continue
lineimage = image.crop(bbox)
lineimage.save(args.pattern % lcount)
f = codecs.open(tpattern % lcount, 'w', 'utf-8')
f.write(get_text(line))
f.close
lcount += 1