Spaces:
Sleeping
Sleeping
File size: 9,295 Bytes
7d30f01 696d188 d3ec064 7d30f01 1867570 a1e81ec b97ae58 de1f3dc b97ae58 f65650a 7d30f01 2ed1ac5 9ec9085 |
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 |
import cv2
import numpy as np
import pytesseract
import tensorflow as tf
from tensorflow.keras.preprocessing.image import img_to_array, load_img
from process import preprocess_image
from PIL import Image
from codecs import encode, decode
import requests
def infer_text(im):
im.save("converted.png")
url = "https://ajax.thehive.ai/api/demo/classify?endpoint=text_recognition"
files = {
"image": ("converted.png", open("converted.png", "rb"), "image/png"),
"model_type": (None, "detection"),
"media_type": (None, "photo"),
}
headers = {"referer": "https://thehive.ai/"}
res = requests.post(url, headers=headers, files=files)
text = ""
for output in res.json()["response"]["output"]:
text += output["block_text"]
text = decode(encode(text, "latin-1", "backslashreplace"), "unicode-escape")
return text
def find_order_id(uploaded_file, input_file, model, ocre):
if ocre == 'Hive':
uploaded_image = Image.open(uploaded_file)
text = infer_text(uploaded_image)
else:
rotated = preprocess_image(uploaded_file)
text = pytesseract.image_to_string(rotated)
with input_file as file:
file_contents = file.read().decode()
lines = file_contents.split('\n')
found = False
possible_order_ids = []
for line in lines:
order_id, name, font = line.strip().split(',')
if name.strip() in text:
image = load_img(uploaded_file, target_size=(64, 64))
image = img_to_array(image)
image = np.expand_dims(image, axis=0)
image = image / 255.0
prediction = model.predict(image)
font_type = 'Pacifico' if prediction[0, 0] > prediction[0, 1] else 'OpenSans-Light'
if font_type == font.strip():
result = {
'status': 'success',
'message': f'Detected Text: {text.strip()}\n, Order ID: {order_id}, Predicted Font Type: {font_type}'
}
found = True
break
else:
possible_order_ids.append(order_id)
if not found:
image = load_img(uploaded_file, target_size=(64, 64))
image = img_to_array(image)
image = np.expand_dims(image, axis=0)
image = image / 255.0
prediction = model.predict(image)
font_type = 'Pacifico' if prediction[0, 0] > prediction[0, 1] else 'OpenSans-Light'
for line in lines:
order_id, name, font = line.strip().split(',')
if font.strip() == font_type:
possible_order_ids.append(order_id)
if len(possible_order_ids) > 0:
result = {
'status': 'warning',
'message': f'Detected Text: {text.strip()}\n, Possible Order IDs: {",".join(possible_order_ids)}, Predicted Font Type: {font_type}'
}
else:
result = {
'status': 'error',
'message': f'Detected Text: {text.strip()}\n, Could not find the Order ID and possible font matches.'
}
return result
def jaccard_similarity(s1, s2):
set1 = set(s1.split())
set2 = set(s2.split())
intersection = len(set1.intersection(set2))
union = len(set1.union(set2))
return intersection / union
def find_order_id_similarity(uploaded_file, input_file, similarity_method, ocre):
if ocre == 'Hive':
uploaded_image = Image.open(uploaded_file)
text = infer_text(uploaded_image)
else:
rotated = preprocess_image(uploaded_file)
text = pytesseract.image_to_string(rotated)
with input_file as file:
file_contents = file.read().decode()
lines = file_contents.split('\n')
if similarity_method == 'exact_match':
for line in lines:
order_id, name, font = line.strip().split(',')
if name.strip() == text.strip():
result = {
'status': 'success',
'message': f'Detected Text: {text.strip()}\n, Order ID: {order_id}'
}
return result
message = f'Detected Text: {text.strip()}\n, Could not find the Order ID.'
result = {'status': 'error', 'message': message}
return result
elif similarity_method == 'jaccard_similarity':
possible_order_ids = []
for line in lines:
order_id, name, font = line.strip().split(',')
jaccard_score = jaccard_similarity(name.strip(), text.strip())
if jaccard_score >= 0.8:
result = {
'status': 'success',
'message': f'Detected Text: {text.strip()}\n, Order ID: {order_id}'
}
return result
elif jaccard_score >= 0.5:
possible_order_ids.append(order_id)
if len(possible_order_ids) > 0:
message = f'Detected Text: {text.strip()}\n, Possible Order IDs: {",".join(possible_order_ids)}'
result = {'status': 'warning', 'message': message}
return result
else:
message = f'Detected Text: {text.strip()}\n, Could not find the Order ID.'
result = {'status': 'error', 'message': message}
return result
def find_order_id_2(uploaded_file, input_file, model, ocre):
if ocre == 'Hive':
uploaded_image = Image.open(uploaded_file)
text = infer_text(uploaded_image)
else:
rotated = preprocess_image(uploaded_file)
text = pytesseract.image_to_string(rotated)
with input_file as file:
file_contents = file.read().decode()
lines = file_contents.split('\n')
found = False
possible_order_ids = []
for line in lines:
order_id, name, font = line.strip().split(',')
if name.strip() in text:
image = load_img(uploaded_file, target_size=(64, 64))
image = img_to_array(image)
image = np.expand_dims(image, axis=0)
image = image / 255.0
prediction = model.predict(image)
class_names = ['Allibretto1.8.otf', 'Bella1.1.otf', 'Buffalo Nickel1.2.otf', 'Cervanttis1.18.otf', 'Claster1.6.otf', 'Fairy4.5.otf', 'Mon-Amour-April1.7.otf', 'Mon-Amour-Aug1.1.otf', 'Mon-Amour-Dec1.2.otf', 'Mon-Amour-Feb1.1.otf', 'Mon-Amour-January1.2.otf', 'Mon-Amour-July1.1.otf', 'Mon-Amour-June1.1.otf', 'Mon-Amour-Mar1.2.otf', 'Mon-Amour-May1.1.otf', 'Mon-Amour-Nov1.1.otf', 'Mon-Amour-Oct1.1.otf', 'Mon-Amour-Sept1.1.otf', 'Mon-Amour2.3.otf', 'Shelby1.3.otf', 'UKIJJ-Quill1.7.otf']
predicted_class_index = np.argmax(prediction[0])
predicted_class_name = class_names[predicted_class_index]
if predicted_class_name.strip() == font.strip():
result = {
'status': 'success',
'message': f'Detected Text: {text.strip()}\n, Order ID: {order_id}, Predicted Font Type: {predicted_class_name.strip()}'
}
found = True
break
else:
possible_order_ids.append(order_id)
if not found:
image = load_img(uploaded_file, target_size=(64, 64))
image = img_to_array(image)
image = np.expand_dims(image, axis=0)
image = image / 255.0
prediction = model.predict(image)
class_names = ['Allibretto1.8.otf', 'Bella1.1.otf', 'Buffalo Nickel1.2.otf', 'Cervanttis1.18.otf', 'Claster1.6.otf', 'Fairy4.5.otf', 'Mon-Amour-April1.7.otf', 'Mon-Amour-Aug1.1.otf', 'Mon-Amour-Dec1.2.otf', 'Mon-Amour-Feb1.1.otf', 'Mon-Amour-January1.2.otf', 'Mon-Amour-July1.1.otf', 'Mon-Amour-June1.1.otf', 'Mon-Amour-Mar1.2.otf', 'Mon-Amour-May1.1.otf', 'Mon-Amour-Nov1.1.otf', 'Mon-Amour-Oct1.1.otf', 'Mon-Amour-Sept1.1.otf', 'Mon-Amour2.3.otf', 'Shelby1.3.otf', 'UKIJJ-Quill1.7.otf']
predicted_class_index = np.argmax(prediction[0])
predicted_class_name = class_names[predicted_class_index]
for line in lines:
order_id, name, font = line.strip().split(',')
if font.strip() == predicted_class_name.strip():
possible_order_ids.append(order_id)
if len(possible_order_ids) > 0:
result = {
'status': 'warning',
'message': f'Detected Text: {text.strip()}\n, Possible Order IDs: {",".join(possible_order_ids)}, Predicted Font Type: {predicted_class_name.strip()}'
}
else:
result = {
'status': 'error',
'message': f'Detected Text: {text.strip()}\n, Could not find the Order ID and possible font matches.'
}
return result |