def recognise_char_sep_model(self, img):
"""Recognition using character model"""
# Pre-processing the word
img = word_normalization(
img,
60,
border=False,
tilt=True,
hyst_norm=True)
# Separate letters
img = cv2.copyMakeBorder(
img,
0, 0, 30, 30,
cv2.BORDER_CONSTANT,
value=[0, 0, 0])
gaps = characters.segment(img, RNN=True)
chars = []
for i in range(len(gaps) - 1):
char = img[:, gaps[i]:gaps[i + 1]]
char, dim = letter_normalization(char, is_thresh=True, dim=True)
# TODO Test different values
if dim[0] > 4 and dim[1] > 4:
chars.append(char.flatten())
chars = np.array(chars)
word = ''
if len(chars) != 0:
pred = CHARACTER_MODEL.run(chars)
for c in pred:
word += idx2char(c)
return word
def words_norm(location, output):
output = os.path.join(location, output)
if not os.path.exists(output):
os.makedirs(output)
else:
print("THIS DATASET IS BEING SKIPPED")
print("Output folder already exists:", output)
return 1
imgs = glob.glob(os.path.join(location, data_folder, '*.png'))
length = len(imgs)
for i, img_path in enumerate(imgs):
image = cv2.imread(img_path)
# Simple check for invalid images
if image.shape[0] > 20:
cv2.imwrite(
os.path.join(output, os.path.basename(img_path)),
word_normalization(image,
height=64,
border=False,
tilt=False,
hyst_norm=False))
print(i)
print_progress_bar(i, len(imgs))
print("\tNumber of normalized words:",
len([n for n in os.listdir(output)]))
def recognise(img):
"""Recognising words using CTC Model."""
img = word_normalization(img,
64,
border=False,
tilt=False,
hyst_norm=False)
length = img.shape[1]
# Input has shape [batch_size, height, width, 1]
input_imgs = np.zeros((1, 64, length, 1), dtype=np.uint8)
input_imgs[0][:, :length, 0] = img
pred = CTC_MODEL.eval_feed({
'inputs:0': input_imgs,
'inputs_length:0': [length],
'keep_prob:0': 1
})[0]
word = ''
for i in pred:
word += idx2char(i)
return word