def test_orderedgroup_export_order():
"""
See https://github.com/OCR-D/core/issues/475
"""
# arrange
with open(assets.path_to('gutachten/data/TEMP1/PAGE_TEMP1.xml'), 'r') as f:
pcgts = parseString(f.read().encode('utf8'), silence=True)
# act
og = pcgts.get_Page().get_ReadingOrder().get_OrderedGroup()
xml_before = to_xml(og)
children = og.get_AllIndexed()
# assert
assert len(children) == 22
assert [c.index for c in children] == list(range(0, 22))
# mix up the indexes
children[0].index = 11
children[11].index = 3
children[3].index = 0
assert [c.index for c in children] == [11, 1, 2, 0, 4, 5, 6, 7, 8, 9, 10, 3, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21]
assert [c.index for c in og.get_AllIndexed()] == list(range(0, 22))
assert og.get_AllIndexed()[1].__class__ == OrderedGroupIndexedType
# serialize and make sure the correct order was serialized
new_pcgts = parseString(to_xml(pcgts).encode('utf8'), silence=True)
new_og = new_pcgts.get_Page().get_ReadingOrder().get_OrderedGroup()
assert [c.index for c in new_og.get_AllIndexed()] == list(range(0, 22))
xml_after = to_xml(new_og)
def test_to_xml(self):
# with open('/tmp/test.xml', 'w') as f:
# f.write(to_xml(self.pcgts))
self.assertIn(
' xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://schema.primaresearch.org/PAGE/gts/pagecontent/2019-07-15 http://schema.primaresearch.org/PAGE/gts/pagecontent/2019-07-15/pagecontent.xsd"',
to_xml(self.pcgts)[:1000])
self.assertIn('</TextRegion', to_xml(self.pcgts))
def process(self):
LOG = getLogger('ocrd.dummy')
assert_file_grp_cardinality(self.input_file_grp, 1)
assert_file_grp_cardinality(self.output_file_grp, 1)
for input_file in self.input_files:
input_file = self.workspace.download_file(input_file)
file_id = make_file_id(input_file, self.output_file_grp)
ext = MIME_TO_EXT.get(input_file.mimetype, '')
local_filename = join(self.output_file_grp, file_id + ext)
pcgts = page_from_file(self.workspace.download_file(input_file))
pcgts.set_pcGtsId(file_id)
self.add_metadata(pcgts)
LOG.info("cp %s %s # %s -> %s", input_file.url, local_filename, input_file.ID, file_id)
if input_file.mimetype == MIMETYPE_PAGE:
# Source file is PAGE-XML: Write out in-memory PcGtsType
self.workspace.add_file(
ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=input_file.mimetype,
local_filename=local_filename,
content=to_xml(pcgts).encode('utf-8'))
else:
# Source file is not PAGE-XML: Copy byte-by-byte
with open(input_file.local_filename, 'rb') as f:
content = f.read()
self.workspace.add_file(
ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=input_file.mimetype,
local_filename=local_filename,
content=content)
if input_file.mimetype.startswith('image/'):
# write out the PAGE-XML representation for this image
page_file_id = file_id + '_PAGE'
pcgts.set_pcGtsId(page_file_id)
pcgts.get_Page().set_imageFilename(local_filename)
page_filename = join(self.output_file_grp, file_id + '.xml')
LOG.info("Add PAGE-XML %s generated for %s at %s",
page_file_id, file_id, page_filename)
self.workspace.add_file(
ID=page_file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=page_filename,
content=to_xml(pcgts).encode('utf-8'))
def process(self):
"""
Performs the line segmentation.
"""
with PyTessBaseAPI(
psm=PSM.SINGLE_LINE,
path=TESSDATA_PREFIX,
) as tessapi:
for (n, input_file) in enumerate(self.input_files):
pcgts = page_from_file(self.workspace.download_file(input_file))
image_url = pcgts.get_Page().imageFilename
for region in pcgts.get_Page().get_TextRegion():
for line in region.get_TextLine():
log.debug("Detecting words in line '%s'", line.id)
image = self.workspace.resolve_image_as_pil(image_url, polygon_from_points(line.get_Coords().points))
tessapi.SetImage(image)
offset = xywh_from_points(line.get_Coords().points)
for (word_no, component) in enumerate(tessapi.GetComponentImages(RIL.WORD, True)):
word_id = '%s_word%04d' % (line.id, word_no)
word_xywh = component[1]
word_xywh['x'] += offset['x']
word_xywh['y'] += offset['y']
line.add_Word(WordType(id=word_id, Coords=CoordsType(points_from_xywh(word_xywh))))
ID = concat_padded(self.output_file_grp, n)
self.workspace.add_file(
ID=ID,
file_grp=self.output_file_grp,
local_filename='%s/%s' % (self.output_file_grp, ID),
mimetype=MIMETYPE_PAGE,
content=to_xml(pcgts).encode('utf-8'),
)
def process(self):
"""
Performs the line segmentation.
"""
with PyTessBaseAPI(path=TESSDATA_PREFIX) as tessapi:
for (n, input_file) in enumerate(self.input_files):
pcgts = page_from_file(
self.workspace.download_file(input_file))
image_url = pcgts.get_Page().imageFilename
for region in pcgts.get_Page().get_TextRegion():
log.debug("Detecting lines in %s with tesseract",
region.id)
image = self.workspace.resolve_image_as_pil(
image_url,
polygon_from_points(region.get_Coords().points))
tessapi.SetImage(image)
offset = xywh_from_points(region.get_Coords().points)
for (line_no, component) in enumerate(
tessapi.GetComponentImages(RIL.TEXTLINE, True)):
line_id = '%s_line%04d' % (region.id, line_no)
line_xywh = component[1]
line_xywh['x'] += offset['x']
line_xywh['y'] += offset['y']
line_points = points_from_xywh(line_xywh)
region.add_TextLine(
TextLineType(id=line_id,
Coords=CoordsType(line_points)))
ID = concat_padded(self.output_file_grp, n)
self.workspace.add_file(
ID=ID,
file_grp=self.output_file_grp,
mimetype=MIMETYPE_PAGE,
local_filename='%s/%s' % (self.output_file_grp, ID),
content=to_xml(pcgts).encode('utf-8'),
)
def process(self):
try:
self.page_grp, self.image_grp = self.output_file_grp.split(',')
except ValueError:
self.page_grp = self.output_file_grp
self.image_grp = FALLBACK_IMAGE_GRP
LOG.info(
"No output file group for images specified, falling back to '%s'",
FALLBACK_IMAGE_GRP)
oplevel = self.parameter['operation_level']
for (n, input_file) in enumerate(self.input_files):
page_id = input_file.pageId or input_file.ID
LOG.info("INPUT FILE %i / %s", n, page_id)
pcgts = page_from_file(self.workspace.download_file(input_file))
metadata = pcgts.get_Metadata()
metadata.add_MetadataItem(
MetadataItemType(
type_="processingStep",
name=self.ocrd_tool['steps'][0],
value=TOOL,
Labels=[
LabelsType( #externalRef="parameters",
Label=[
LabelType(type_=name,
value=self.parameter[name])
for name in self.parameter.keys()
])
]))
page = pcgts.get_Page()
angle = page.get_orientation()
if angle:
LOG.warning('Overwriting existing deskewing angle: %i', angle)
page_image, page_xywh, page_image_info = self.workspace.image_from_page(
page, page_id, feature_filter='deskewed')
if oplevel == "page":
self._process_segment(page_image, page, page_xywh, page_id,
input_file, n)
else:
LOG.warning('Operation level %s, but should be "page".',
oplevel)
break
# Use input_file's basename for the new file -
# this way the files retain the same basenames:
file_id = input_file.ID.replace(self.input_file_grp,
self.output_file_grp)
if file_id == input_file.ID:
file_id = concat_padded(self.output_file_grp, n)
self.workspace.add_file(ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(
self.output_file_grp,
file_id + '.xml'),
content=to_xml(pcgts).encode('utf-8'))
def process(self):
"""
Segment with ocropy
"""
for (n, input_file) in enumerate(self.input_files):
log.info("INPUT FILE %i / %s", n, input_file)
downloaded_file = self.workspace.download_file(input_file)
log.info("downloaded_file %s", downloaded_file)
pcgts = page_from_file(downloaded_file)
page_width = pcgts.get_Page().get_imageWidth()
page_height = pcgts.get_Page().get_imageHeight()
# TODO binarized variant from get_AlternativeImage()
image_url = pcgts.get_Page().imageFilename
log.info("pcgts %s", pcgts)
binary = ocrolib.read_image_binary(
self.workspace.download_url(image_url))
binary = 1 - binary
scale = self.parameter['scale'] if self.parameter[
'scale'] != 0 else psegutils.estimate_scale(binary)
log.debug(binary)
pseg = self.compute_segmentation(binary, scale)
log.debug("pseg=%s", pseg)
# TODO reading order / enumber
# log.debug("finding reading order")
# lines = psegutils.compute_lines(pseg, scale)
# order = psegutils.reading_order([l.bounds for l in lines])
# lsort = psegutils.topsort(order)
regions = ocrolib.RegionExtractor()
regions.setPageLines(pseg)
dummyRegion = TextRegionType(
id="dummy",
Coords=CoordsType(
points="0,0 %s,0 %s,%s 0,%s" %
(page_width, page_width, page_height, page_height)))
pcgts.get_Page().add_TextRegion(dummyRegion)
for lineno in range(1, regions.length()):
log.debug("id=%s bbox=%s", regions.id(lineno),
regions.bbox(lineno))
textline = TextLineType(
id=concat_padded("line", lineno),
Coords=CoordsType(
points=points_from_y0x0y1x1(regions.bbox(lineno))))
dummyRegion.add_TextLine(textline)
ID = concat_padded(self.output_file_grp, n)
self.workspace.add_file(ID=ID,
file_grp=self.output_file_grp,
mimetype=MIMETYPE_PAGE,
local_filename="%s/%s.xml" %
(self.output_file_grp, ID),
content=to_xml(pcgts))
def process(self):
network_file = self.parameter['network']
stride = self.parameter['stride']
classifier = TypegroupsClassifier.load(network_file)
ignore_type = ('Adornment', 'Book covers and other irrelevant data',
'Empty Pages', 'Woodcuts - Engravings')
self.log.debug('Processing: %s', self.input_files)
for (_, input_file) in enumerate(self.input_files):
pcgts = page_from_file(self.workspace.download_file(input_file))
image_url = pcgts.get_Page().imageFilename
pil_image = self.workspace.resolve_image_as_pil(image_url)
result = classifier.run(pil_image, stride)
score_sum = 0
for typegroup in classifier.classMap.cl2id:
if not typegroup in ignore_type:
score_sum += max(0, result[typegroup])
script_highscore = 0
noise_highscore = 0
result_map = {}
output = ''
for typegroup in classifier.classMap.cl2id:
score = result[typegroup]
if typegroup in ignore_type:
noise_highscore = max(noise_highscore, score)
else:
script_highscore = max(script_highscore, score)
normalised_score = max(0, score / score_sum)
result_map[normalised_score] = typegroup
if noise_highscore > script_highscore:
pcgts.get_Page().set_primaryScript(None)
self.log.debug(
'Detected only noise (such as empty page or book cover). noise_highscore=%s > script_highscore=%s',
noise_highscore, script_highscore)
else:
for k in sorted(result_map, reverse=True):
intk = round(100 * k)
if intk <= 0:
continue
if output != '':
output = '%s, ' % output
output = '%s%s:%d' % (output, result_map[k], intk)
self.log.debug('Detected %s' % output)
page = pcgts.get_Page()
textStyle = page.get_TextStyle()
if not textStyle:
textStyle = TextStyleType()
page.set_TextStyle(textStyle)
textStyle.set_fontFamily(output)
ID = concat_padded(self.output_file_grp, input_file.ID)
self.workspace.add_file(ID=ID,
file_grp=self.output_file_grp,
mimetype=MIMETYPE_PAGE,
local_filename="%s/%s" %
(self.output_file_grp, ID),
content=to_xml(pcgts))
def process(self):
"""
Performs the cropping.
"""
with tesserocr.PyTessBaseAPI(path=TESSDATA_PREFIX) as tessapi:
# print(self.input_file_grp)
for (n, input_file) in enumerate(self.input_files):
# print(input_file)
pcgts = page_from_file(self.workspace.download_file(input_file))
image = self.workspace.resolve_image_as_pil(pcgts.get_Page().imageFilename)
log.debug("Cropping with tesseract")
tessapi.SetImage(image)
#
# helper variables for saving the box coordinates
#
min_x = image.width
min_y = image.height
max_x = 0
max_y = 0
# iterate over all boxes and compare their extent
# to the min and max values
for component in tessapi.GetComponentImages(tesserocr.RIL.BLOCK, True):
points, index = points_from_xywh(component[1]), component[2]
#
# the region reference in the reading order element
#
ID = "region%04d" % index
log.debug("Detected region '%s': %s", ID, points)
for pair in points.split(' '):
x, y = (int(pair.split(',')[0]), int(pair.split(',')[1]))
if x < min_x:
min_x = x
if y < min_y:
min_y = y
elif x > max_x:
max_x = x
elif y > max_y:
max_y = y
log.debug("Updated page border: %i,%i %i,%i %i,%i %i,%i" % (min_x, min_y, max_x, min_y, max_x, max_y, min_x, max_y))
#
# set the identified page border
#
brd = BorderType(Coords=CoordsType("%i,%i %i,%i %i,%i %i,%i" % (min_x, min_y, max_x, min_y, max_x, max_y, min_x, max_y)))
pcgts.get_Page().set_Border(brd)
ID = concat_padded(self.output_file_grp, n)
self.workspace.add_file(
ID=ID,
file_grp=self.output_file_grp,
mimetype=MIMETYPE_PAGE,
local_filename='%s/%s' % (self.output_file_grp, ID),
content=to_xml(pcgts).encode('utf-8'),
)
def test_serialize_no_empty_readingorder():
"""
https://github.com/OCR-D/core/issues/602
"""
pcgts = page_from_image(create_ocrd_file_with_defaults(url=assets.path_to('kant_aufklaerung_1784/data/OCR-D-IMG/INPUT_0017.tif')))
pcgts.get_Page().set_ReadingOrder(ReadingOrderType())
assert pcgts.get_Page().get_ReadingOrder()
pcgts = parseString(to_xml(pcgts, skip_declaration=True))
assert not pcgts.get_Page().get_ReadingOrder()
def test_to_xml_unicode_nsprefix(self):
with open(assets.path_to('kant_aufklaerung_1784-binarized/data/OCR-D-GT-WORD/INPUT_0020.xml'), 'rb') as f:
from_xml = f.read()
self.assertIn('<Unicode>', from_xml.decode('utf-8'), 'without NS prefix')
self.assertIn('<Created', from_xml.decode('utf-8'), 'without NS prefix')
pcgts = parseString(from_xml, silence=True)
as_xml = to_xml(pcgts)
self.assertIn('<pc:Unicode>', as_xml, 'with NS prefix')
self.assertIn('<pc:Created>', as_xml, 'with NS prefix')
def process(self):
"""
Performs the (text) recognition.
"""
# print(self.parameter)
self.maxlevel = self.parameter['textequiv_level']
linesdir = self.parameter['linesdir']
if self.maxlevel not in ['line', 'word', 'glyph']:
raise Exception(
"currently only implemented at the line/glyph level")
root, _, files = os.walk(linesdir).__next__()
self.root = root
predfiles = []
for file in files:
if '.pred' in file:
predfiles.append(file[:-9])
########################################################################################
# self.log.info("Using model %s in %s for recognition", model)
for (n, input_file) in enumerate(self.input_files):
# self.log.info("INPUT FILE %i / %s", n, input_file)
pcgts = page_from_file(self.workspace.download_file(input_file))
self.log.info("Processing text in page '%s'", pcgts.get_pcGtsId())
page = pcgts.get_Page()
index = input_file.url.rfind('/') + 1
fgrp = input_file.url[index:-4]
# region, line, word, or glyph level:
regions = page.get_TextRegion()
if not regions:
self.log.warning("Page contains no text regions")
self.process_regions(regions, predfiles, fgrp)
ID = concat_padded(self.output_file_grp, n)
self.log.info('creating file id: %s, name: %s, file_grp: %s', ID,
input_file.basename, self.output_file_grp)
# Use the input file's basename for the new file
# this way the files retain the same basenames.
out = self.workspace.add_file(
ID=ID,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
basename=self.output_file_grp + '-' + input_file.basename,
mimetype=MIMETYPE_PAGE,
content=to_xml(pcgts),
)
self.log.info('created file %s', out)
def redraw(self) -> None:
if self.current:
self.text_view.set_tooltip_text(self.page_id)
if self.current.file:
with self.document.path(self.current.file).open('r') as f:
text = f.read()
else:
text = to_xml(self.current.pc_gts)
self.buffer.set_text(text)
else:
self.buffer.set_text('')
def process(self):
try:
self.page_grp, self.image_grp = self.output_file_grp.split(',')
except ValueError:
self.page_grp = self.output_file_grp
self.image_grp = FALLBACK_IMAGE_GRP
LOG.info(
"No output file group for images specified, falling back to '%s'",
FALLBACK_IMAGE_GRP)
oplevel = self.parameter['operation_level']
for (n, input_file) in enumerate(self.input_files):
file_id = input_file.ID.replace(self.input_file_grp,
self.image_grp)
page_id = input_file.pageId or input_file.ID
LOG.info("INPUT FILE %i / %s", n, page_id)
pcgts = page_from_file(self.workspace.download_file(input_file))
metadata = pcgts.get_Metadata()
metadata.add_MetadataItem(
MetadataItemType(
type_="processingStep",
name=self.ocrd_tool['steps'][0],
value=TOOL,
Labels=[
LabelsType( #externalRef="parameters",
Label=[
LabelType(type_=name,
value=self.parameter[name])
for name in self.parameter.keys()
])
]))
page = pcgts.get_Page()
page_image, page_xywh, page_image_info = self.workspace.image_from_page(
page, page_id)
if oplevel == "page":
self._process_segment(page, page_image.filename, page_id,
file_id + ".ds")
file_id = input_file.ID.replace(self.input_file_grp,
self.output_file_grp)
if file_id == input_file.ID:
file_id = concat_padded(self.output_file_grp, n)
self.workspace.add_file(ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(
self.output_file_grp,
file_id + '.xml'),
content=to_xml(pcgts).encode('utf-8'))
def process(self):
"""Segment pages into regions using a Mask R-CNN model."""
assert_file_grp_cardinality(self.input_file_grp, 1)
assert_file_grp_cardinality(self.output_file_grp, 1)
LOG = getLogger('processor.AnybaseocrBlockSegmenter')
if not tf.test.is_gpu_available():
LOG.warning(
"Tensorflow cannot detect CUDA installation. Running without GPU will be slow."
)
for input_file in self.input_files:
pcgts = page_from_file(self.workspace.download_file(input_file))
self.add_metadata(pcgts)
page = pcgts.get_Page()
page_id = input_file.pageId or input_file.ID
# todo rs: why not cropped?
page_image, page_xywh, page_image_info = self.workspace.image_from_page(
page,
page_id,
feature_filter='binarized,deskewed,cropped,clipped,non_text')
# try to load pixel masks
try:
# todo rs: this combination only works for tiseg with use_deeplr=true
mask_image, _, _ = self.workspace.image_from_page(
page,
page_id,
feature_selector='clipped',
feature_filter='binarized,deskewed,cropped,non_text')
except:
mask_image = None
if page_image_info.resolution != 1:
dpi = page_image_info.resolution
if page_image_info.resolutionUnit == 'cm':
dpi = round(dpi * 2.54)
else:
dpi = None
self._process_segment(page_image, page, page_xywh, page_id,
input_file, mask_image, dpi)
file_id = make_file_id(input_file, self.output_file_grp)
pcgts.set_pcGtsId(file_id)
self.workspace.add_file(ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(
self.output_file_grp,
file_id + '.xml'),
content=to_xml(pcgts).encode('utf-8'))
def test_simpletypes(self):
pcgts = parseString(simple_page, silence=True)
self.assertTrue(isinstance(pcgts.get_Page().imageWidth, int))
el = pcgts.get_Page().get_TextRegion()[0].get_TextLine()[0].get_Word(
)[0].get_TextEquiv()[0]
self.assertTrue(isinstance(el.conf, float))
# XXX no validation on setting attributes :-(
# c.f. https://www.davekuhlman.org/generateDS.html#simpletype
# el.set_conf('2.0987')
# self.assertTrue(isinstance(el.conf, float))
with self.assertRaisesRegex(TypeError, ''):
el.set_conf('I AM NOT A FLOAT DEAL WITH IT')
parseString(to_xml(pcgts).encode('utf8'))
def process(self):
"""
Performs the (text) recognition.
"""
mainIndex = self.parameter['mainIndex']
for (n, input_file) in enumerate(self.input_files):
alignurl = input_file.url
pcgts = parse(alignurl, True)
page = pcgts.get_Page()
regions = page.get_TextRegion()
pagecontent = ''
for region in regions:
regioncontent = ''
lines = region.get_TextLine()
for line in lines:
linecontent = ''
words = line.get_Word()
for word in words:
wordunicode = word.get_TextEquiv()[mainIndex].Unicode
word.add_TextEquiv(TextEquivType(Unicode=wordunicode))
linecontent += ' ' + wordunicode
line.add_TextEquiv(TextEquivType(Unicode=regioncontent))
regioncontent += '\n' + linecontent
region.add_TextEquiv(TextEquivType(Unicode=regioncontent))
pagecontent += '\n' + regioncontent
page.add_TextEquiv(TextEquivType(Unicode=pagecontent))
ID = concat_padded(self.output_file_grp, n)
self.log.info('creating file id: %s, name: %s, file_grp: %s', ID,
input_file.basename, self.output_file_grp)
# Use the input file's basename for the new file
# this way the files retain the same basenames.
out = self.workspace.add_file(
ID=ID,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
basename=self.output_file_grp + '-' + input_file.basename,
mimetype=MIMETYPE_PAGE,
content=to_xml(pcgts),
)
self.log.info('created file %s', out)
def test_to_xml_unicode_nsprefix():
"""see https://github.com/OCR-D/core/pull/474#issuecomment-621477590"""
# arrange
with open(assets.path_to('kant_aufklaerung_1784-binarized/data/OCR-D-GT-WORD/INPUT_0020.xml'), 'rb') as f:
from_xml = f.read()
# assert
assert '<Unicode>' in from_xml.decode('utf-8'), 'without NS prefix'
assert '<Created' in from_xml.decode('utf-8'), 'without NS prefix'
pcgts = parseString(from_xml, silence=True)
as_xml = to_xml(pcgts)
assert '<pc:Unicode>' in as_xml, 'with NS prefix'
assert '<pc:Created>' in as_xml, 'with NS prefix'
def process(self):
"""
Performs the recognition.
"""
self._init_calamari()
for (n, input_file) in enumerate(self.input_files):
page_id = input_file.pageId or input_file.ID
log.info("INPUT FILE %i / %s", n, page_id)
pcgts = page_from_file(self.workspace.download_file(input_file))
page = pcgts.get_Page()
page_image, page_xywh, page_image_info = self.workspace.image_from_page(page, page_id)
for region in pcgts.get_Page().get_TextRegion():
region_image, region_xywh = self.workspace.image_from_segment(region, page_image, page_xywh)
textlines = region.get_TextLine()
log.info("About to recognize %i lines of region '%s'", len(textlines), region.id)
for (line_no, line) in enumerate(textlines):
log.debug("Recognizing line '%s' in region '%s'", line_no, region.id)
line_image, line_xywh = self.workspace.image_from_segment(line, region_image, region_xywh)
line_image_np = np.array(line_image, dtype=np.uint8)
raw_results = list(self.predictor.predict_raw([line_image_np], progress_bar=False))[0]
for i, p in enumerate(raw_results):
p.prediction.id = "fold_{}".format(i)
prediction = self.voter.vote_prediction_result(raw_results)
prediction.id = "voted"
line_text = prediction.sentence
line_conf = prediction.avg_char_probability
line.set_TextEquiv([TextEquivType(Unicode=line_text, conf=line_conf)])
_page_update_higher_textequiv_levels('line', pcgts)
file_id = self._make_file_id(input_file, n)
self.workspace.add_file(
ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(self.output_file_grp, file_id + '.xml'),
content=to_xml(pcgts))
def process(self):
LOG = getLogger('OcrdAnybaseocrTextline')
assert_file_grp_cardinality(self.input_file_grp, 1)
assert_file_grp_cardinality(self.output_file_grp, 1)
oplevel = self.parameter['operation_level']
for (n, input_file) in enumerate(self.input_files):
page_id = input_file.pageId or input_file.ID
pcgts = page_from_file(self.workspace.download_file(input_file))
self.add_metadata(pcgts)
page = pcgts.get_Page()
LOG.info("INPUT FILE %s", input_file.pageId or input_file.ID)
page_image, page_xywh, page_image_info = self.workspace.image_from_page(
page, page_id, feature_selector='binarized,deskewed')
if oplevel == 'page':
LOG.warning("Operation level should be region.")
self._process_segment(page_image, page, None, page_xywh,
page_id, input_file, n)
else:
regions = page.get_TextRegion()
if not regions:
LOG.warning("Page '%s' contains no text regions", page_id)
continue
for (k, region) in enumerate(regions):
region_image, region_xywh = self.workspace.image_from_segment(
region, page_image, page_xywh)
self._process_segment(region_image, page, region,
region_xywh, region.id, input_file,
k)
file_id = make_file_id(input_file, self.output_file_grp)
pcgts.set_pcGtsId(file_id)
self.workspace.add_file(ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(
self.output_file_grp,
file_id + '.xml'),
content=to_xml(pcgts).encode('utf-8'))
def negative2zero(inputfile, outputfile):
print("Setting negative coords to zero..")
pcgts = parse(inputfile, silence=True)
page = pcgts.get_Page()
for attr in dir(page):
if "get_" in attr and "Region" in attr:
for regiondata in getattr(page,attr)():
if attr == "get_TextRegion":
for textline in regiondata.get_TextLine():
textcoords = textline.get_Coords()
textcoords.set_points(update_points(textcoords.get_points()))
regcoords = regiondata.get_Coords()
regcoords.set_points(update_points(regcoords.get_points()))
content = to_xml(pcgts)
with open(outputfile,"w") as fout:
fout.write(content)
def process(self):
"""
Performs the region segmentation.
"""
with tesserocr.PyTessBaseAPI(path=TESSDATA_PREFIX) as tessapi:
# print(self.input_file_grp)
for (n, input_file) in enumerate(self.input_files):
# print(input_file)
pcgts = page_from_file(self.workspace.download_file(input_file))
image = self.workspace.resolve_image_as_pil(pcgts.get_Page().imageFilename)
log.debug("Detecting regions with tesseract")
tessapi.SetImage(image)
for component in tessapi.GetComponentImages(tesserocr.RIL.BLOCK, True):
points, index = points_from_xywh(component[1]), component[2]
#
# the region reference in the reading order element
#
ID = "region%04d" % index
log.debug("Detected region '%s': %s", ID, points)
# <pg:ReadingOrder>
ro = pcgts.get_Page().get_ReadingOrder()
if ro is None:
ro = ReadingOrderType()
pcgts.get_Page().set_ReadingOrder(ro)
# <pg:OrderedGroup>
og = ro.get_OrderedGroup()
if og is None:
og = OrderedGroupType(id="reading-order")
ro.set_OrderedGroup(og)
# <pg:RegionRefIndexed>
og.add_RegionRefIndexed(RegionRefIndexedType(regionRef=ID, index=index))
#
# text region
#
pcgts.get_Page().add_TextRegion(TextRegionType(id=ID, Coords=CoordsType(points=points)))
ID = concat_padded(self.output_file_grp, n)
self.workspace.add_file(
ID=ID,
file_grp=self.output_file_grp,
mimetype=MIMETYPE_PAGE,
local_filename='%s/%s' % (self.output_file_grp, ID),
content=to_xml(pcgts).encode('utf-8'),
)
def process(self):
"""Performs border detection on the workspace. """
assert_file_grp_cardinality(self.input_file_grp, 1)
assert_file_grp_cardinality(self.output_file_grp, 1)
LOG = getLogger('OcrdAnybaseocrCropper')
oplevel = self.parameter['operation_level']
for (n, input_file) in enumerate(self.input_files):
page_id = input_file.pageId or input_file.ID
LOG.info("INPUT FILE %i / %s", n, page_id)
pcgts = page_from_file(self.workspace.download_file(input_file))
self.add_metadata(pcgts)
page = pcgts.get_Page()
# Check for existing Border --> already cropped
border = page.get_Border()
if border:
left, top, right, bottom = bbox_from_points(
border.get_Coords().points)
LOG.warning('Overwriting existing Border: %i:%i,%i:%i',
left, top, right, bottom)
page = pcgts.get_Page()
page_image, page_coords, page_image_info = self.workspace.image_from_page(
page, page_id,
feature_filter='cropped',
feature_selector='binarized') # should also be deskewed
if oplevel == "page":
self._process_segment(
page_image, page, page_coords, page_id, input_file, n)
else:
raise Exception(
'Operation level %s, but should be "page".', oplevel)
file_id = make_file_id(input_file, self.output_file_grp)
pcgts.set_pcGtsId(file_id)
self.workspace.add_file(
ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(self.output_file_grp, file_id + '.xml'),
content=to_xml(pcgts).encode('utf-8')
)
def process(self):
for (n, input_file) in enumerate(self.input_files):
LOG.info("INPUT FILE %i / %s", n, input_file)
local_input_file = self.workspace.download_file(input_file)
pcgts = parse(local_input_file.url, silence=True)
LOG.info("Scoring text in page '%s' at the %s level",
pcgts.get_pcGtsId(), self.parameter['textequiv_level'])
self._process_page(pcgts)
# write back result
file_id = concat_padded(self.output_file_grp, n)
self.workspace.add_file(
ID=file_id,
file_grp=self.output_file_grp,
local_filename=os.path.join(self.output_file_grp, file_id),
mimetype=MIMETYPE_PAGE,
content=to_xml(pcgts),
)
def process(self):
assert_file_grp_cardinality(self.input_file_grp, 1)
assert_file_grp_cardinality(self.output_file_grp, 1)
oplevel = self.parameter['operation_level']
LOG = getLogger('OcrdAnybaseocrBinarizer')
for (n, input_file) in enumerate(self.input_files):
file_id = make_file_id(input_file, self.output_file_grp)
page_id = input_file.pageId or input_file.ID
LOG.info("INPUT FILE %i / %s", n, page_id)
pcgts = page_from_file(self.workspace.download_file(input_file))
pcgts.set_pcGtsId(file_id)
self.add_metadata(pcgts)
page = pcgts.get_Page()
page_image, page_xywh, page_image_info = self.workspace.image_from_page(
page, page_id, feature_filter="binarized")
LOG.info("Binarizing on '%s' level in page '%s'", oplevel, page_id)
if oplevel == "page":
self._process_segment(page_image, page, page_xywh, page_id,
input_file, n)
else:
regions = page.get_TextRegion() + page.get_TableRegion()
if not regions:
LOG.warning("Page '%s' contains no text regions", page_id)
for (k, region) in enumerate(regions):
region_image, region_xywh = self.workspace.image_from_segment(
region, page_image, page_xywh)
# TODO: not tested on regions
self._process_segment(region_image, page, region_xywh,
region.id, input_file,
str(n) + "_" + str(k))
self.workspace.add_file(ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(
self.output_file_grp,
file_id + '.xml'),
content=to_xml(pcgts).encode('utf-8'))
def process(self):
for (n, input_file) in enumerate(self.input_files):
page_id = input_file.pageId or input_file.ID
LOG.info("INPUT FILE %i / %s", n, page_id)
pcgts = page_from_file(self.workspace.download_file(input_file))
page = pcgts.get_Page()
regions = page.get_TextRegion()
for region in regions:
if region.readingDirection != 'left-to-right':
LOG.info('Not processing region "%s" (not left-to-right)',
region.id)
continue
if len(region.get_TextLine()
) > 1 and region.textLineOrder != 'top-to-bottom':
LOG.info('Not processing region "%s" (not top-to-bottom)',
region.id)
continue
_fix_lines(region)
lines = region.get_TextLine()
for line in lines:
_fix_words(line)
words = line.get_Word()
for word in words:
_fix_glyphs(word)
file_id = input_file.ID.replace(self.input_file_grp,
self.output_file_grp)
if file_id == input_file.ID:
file_id = concat_padded(self.output_file_grp, n)
self.workspace.add_file(ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(
self.output_file_grp,
file_id + '.xml'),
content=to_xml(pcgts))
def test_empty_groups_to_regionrefindexed(self):
"""
Corrolary See https://github.com/OCR-D/core/issues/475
"""
with open(assets.path_to('gutachten/data/TEMP1/PAGE_TEMP1.xml'),
'r') as f:
pcgts = parseString(f.read().encode('utf8'), silence=True)
og = pcgts.get_Page().get_ReadingOrder().get_OrderedGroup()
children = og.get_AllIndexed()
self.assertTrue(isinstance(children[1], OrderedGroupIndexedType))
self.assertTrue(isinstance(children[21],
UnorderedGroupIndexedType))
# empty all the elements in the first orederdGroupIndexed
children[1].set_RegionRefIndexed([])
# serialize apnd parse to see empty group converted
pcgts = parseString(to_xml(pcgts).encode('utf8'), silence=True)
og = pcgts.get_Page().get_ReadingOrder().get_OrderedGroup()
children = og.get_AllIndexed()
self.assertTrue(isinstance(children[1], RegionRefIndexedType))
self.assertTrue(isinstance(children[21], RegionRefIndexedType))
def process(self):
assert_file_grp_cardinality(self.input_file_grp, 1)
assert_file_grp_cardinality(self.output_file_grp, 1)
oplevel = self.parameter['operation_level']
LOG = getLogger('OcrdAnybaseocrDeskewer')
for (n, input_file) in enumerate(self.input_files):
page_id = input_file.pageId or input_file.ID
LOG.info("INPUT FILE %i / %s", n, page_id)
pcgts = page_from_file(self.workspace.download_file(input_file))
self.add_metadata(pcgts)
page = pcgts.get_Page()
angle = page.get_orientation()
if angle:
LOG.warning('Overwriting existing deskewing angle: %i', angle)
page_image, page_xywh, page_image_info = self.workspace.image_from_page(
page,
page_id,
feature_filter='deskewed',
feature_selector='binarized')
if oplevel == "page":
self._process_segment(page_image, page, page_xywh, page_id,
input_file, n)
else:
LOG.warning('Operation level %s, but should be "page".',
oplevel)
break
file_id = make_file_id(input_file, self.output_file_grp)
pcgts.set_pcGtsId(file_id)
self.workspace.add_file(ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
mimetype=MIMETYPE_PAGE,
local_filename=os.path.join(
self.output_file_grp,
file_id + '.xml'),
content=to_xml(pcgts).encode('utf-8'))
def process(self):
"""
Segment with kraken
"""
log = getLogger('processor.KrakenSegment')
for (n, input_file) in enumerate(self.input_files):
log.info("INPUT FILE %i / %s", n, input_file)
downloaded_file = self.workspace.download_file(input_file)
log.info("downloaded_file %s", downloaded_file)
pcgts = page_from_file(downloaded_file)
# TODO binarized variant from get_AlternativeImage()
image_url = pcgts.get_Page().imageFilename
log.info("pcgts %s", pcgts)
im = self.workspace.resolve_image_as_pil(image_url)
log.info('Segmenting')
log.info('Params %s', self.parameter)
res = segment(im, self.parameter['text_direction'],
self.parameter['scale'],
self.parameter['maxcolseps'],
self.parameter['black_colseps'])
if self.parameter['script_detect']:
res = detect_scripts(im, res)
dummyRegion = TextRegionType()
pcgts.get_Page().add_TextRegion(dummyRegion)
# print(res)
for lineno, box in enumerate(res['boxes']):
textline = TextLineType(
id=concat_padded("line", lineno),
Coords=CoordsType(points=points_from_x0y0x1y1(box)))
dummyRegion.add_TextLine(textline)
ID = concat_padded(self.output_file_grp, n)
self.workspace.add_file(self.output_file_grp,
pageId=input_file.pageId,
ID=ID,
mimetype=MIMETYPE_PAGE,
local_filename="%s/%s.xml" %
(self.output_file_grp, ID),
content=to_xml(pcgts).encode('utf-8'))
def process(self):
ifgs = self.input_file_grp.split(",") # input file groups
if len(ifgs) < 2:
raise Exception("need at least two input file groups to align")
ifts = self.zip_input_files(ifgs) # input file tuples
for _id, ift in enumerate(ifts):
alignments = json.loads(self.run_java_aligner(ift))
pcgts = self.align(alignments, ift)
# keep the right part after OCR-D-...-filename
# and prepend output_file_grp
input_file = ift[0].input_file
file_id = make_file_id(input_file, self.output_file_grp)
pcgts.set_pcGtsId(file_id)
out = self.workspace.add_file(
ID=file_id,
file_grp=self.output_file_grp,
pageId=input_file.pageId,
local_filename=os.path.join(self.output_file_grp, file_id + '.xml'),
mimetype=MIMETYPE_PAGE,
content=to_xml(pcgts),
)
self.log.info('created file %s', out)