def parse_pages(self):
"""Genereate metadata for full page images.
Includes every character on page. Characters not in character
set or vocabulary will be labeled as unknown when converted to
integer IDs.
Returns:
:obj:`list` of :obj:`carpedm.data.util.ImageMeta`:
Page image meta data.
"""
pages = []
char = self.character(next(self._reader))
image_id = char.image_id
image = ImageMeta(filepath=image_id, full_image=True, first_char=char)
for char in self.characters():
if char.image_id == image_id:
image.add_char(char)
else:
pages.append(image)
image_id = char.image_id
image = ImageMeta(filepath=image_id,
full_image=True,
first_char=char)
# Add last image.
pages.append(image)
return pages
def test_character_bounding_boxes(self):
meta = ImageMeta(filepath=self.char1.image_id, full_image=False,
first_char=self.char1)
meta.add_char(self.char2)
# char1
self.assertEqual(meta.char_bboxes[0].xmin, 0)
self.assertEqual(meta.char_bboxes[0].xmax, 179)
self.assertEqual(meta.char_bboxes[0].ymin, 0)
self.assertEqual(meta.char_bboxes[0].ymax, 132)
# char2
self.assertEqual(meta.char_bboxes[1].xmin, 8)
self.assertEqual(meta.char_bboxes[1].xmax, 157)
self.assertEqual(meta.char_bboxes[1].ymin, 246)
self.assertEqual(meta.char_bboxes[1].ymax, 414)
def parse_sequences(self, charset, len_min, len_max):
"""Generate metadata for images of character sequences.
Only includes sequences of chars in the desired character set.
If ``len_min == len_max``, sequence length is deterministic, else
each sequence is of random length from [len_min, len_max].
Args:
charset (CharacterSet): The character set.
len_min (int): Minimum sequence length.
len_max (int): Maximum sequence length.
Returns:
:obj:`list` of :obj:`carpedm.data.util.ImageMeta`:
Sequence image meta data.
"""
sequences = []
length = random.randint(len_min, len_max)
image = None
for c in self.characters():
if image is None:
if charset.in_charset(c.label):
image = ImageMeta(filepath=c.image_id, first_char=c)
elif (image.valid_char(c, same_line=True)
and charset.in_charset(c.label)
and image.num_chars < length):
image.add_char(c)
else:
if len_min <= image.num_chars <= len_max:
sequences.append(image)
length = random.randint(len_min, len_max)
if charset.in_charset(c.label):
image = ImageMeta(filepath=c.image_id, first_char=c)
else:
image = None
if image is not None and len_min <= image.num_chars <= len_max:
sequences.append(image)
return sequences
def parse_lines(self):
"""Generate metadata for vertical lines of characters.
Characters not in character set or vocabulary will be labeled as
unknown when converted to integer IDs.
Returns:
:obj:`list` of :obj:`carpedm.data.util.ImageMeta`:
Line image meta data.
"""
lines = []
c = self.character(next(self._reader))
image = ImageMeta(filepath=c.image_id, first_char=c)
for c in self.characters():
if image.valid_char(c, same_line=True):
image.add_char(c)
else:
lines.append(image)
image = ImageMeta(filepath=c.image_id, first_char=c)
lines.append(image)
return lines
def test_line_bounding_boxes_for_full_page(self):
meta = ImageMeta(filepath=self.image_path, full_image=True)
with open(self.csv_path, 'r') as csvfile:
reader = csv.reader(csvfile)
next(reader) # header
for row in reader:
meta.add_char(Character(*row))
# line1
self.assertEqual(meta.line_bboxes[0].xmin, 2290)
self.assertEqual(meta.line_bboxes[0].xmax, 2493)
self.assertEqual(meta.line_bboxes[0].ymin, 717)
self.assertEqual(meta.line_bboxes[0].ymax, 3780)
# line2
self.assertEqual(meta.line_bboxes[1].xmin, 1931)
self.assertEqual(meta.line_bboxes[1].xmax, 2136)
self.assertEqual(meta.line_bboxes[1].ymin, 703)
self.assertEqual(meta.line_bboxes[1].ymax, 3796)
# line3
self.assertEqual(meta.line_bboxes[2].xmin, 1606)
self.assertEqual(meta.line_bboxes[2].xmax, 1811)
self.assertEqual(meta.line_bboxes[2].ymin, 715)
self.assertEqual(meta.line_bboxes[2].ymax, 3783)
# line4
self.assertEqual(meta.line_bboxes[3].xmin, 1266)
self.assertEqual(meta.line_bboxes[3].xmax, 1485)
self.assertEqual(meta.line_bboxes[3].ymin, 701)
self.assertEqual(meta.line_bboxes[3].ymax, 3320)
# line5
self.assertEqual(meta.line_bboxes[4].xmin, 940)
self.assertEqual(meta.line_bboxes[4].xmax, 1154)
self.assertEqual(meta.line_bboxes[4].ymin, 859)
self.assertEqual(meta.line_bboxes[4].ymax, 2368)
# line6
self.assertEqual(meta.line_bboxes[5].xmin, 612)
self.assertEqual(meta.line_bboxes[5].xmax, 817)
self.assertEqual(meta.line_bboxes[5].ymin, 1627)
self.assertEqual(meta.line_bboxes[5].ymax, 3583)
def test_character_labels(self):
meta = ImageMeta(filepath=self.char1.image_id, full_image=False,
first_char=self.char1)
meta.add_char(self.char2)
self.assertEqual(meta.char_labels, ['U+4E4B', 'U+6CD5'])
def test_num_characters(self):
meta = ImageMeta(filepath=self.char1.image_id, full_image=False,
first_char=self.char1)
meta.add_char(self.char2)
self.assertEqual(meta.num_chars, 2)
def test_get_multi_char_height(self):
meta = ImageMeta(filepath=self.char1.image_id, full_image=False,
first_char=self.char1)
meta.add_char(self.char2)
self.assertEqual(meta.height, 414)
def test_get_multi_char_width(self):
meta = ImageMeta(filepath=self.char1.image_id, full_image=False,
first_char=self.char1)
meta.add_char(self.char2)
self.assertEqual(meta.width, 179)
def test_get_multi_char_ymax(self):
meta = ImageMeta(filepath=self.char1.image_id, full_image=False,
first_char=self.char1)
meta.add_char(self.char2)
self.assertEqual(meta.ymax, 963 + 168)