Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -10,7 +10,8 @@ os.environ['USE_TORCH'] = '1'
|
|
| 10 |
import matplotlib.pyplot as plt
|
| 11 |
from doctr.io import DocumentFile
|
| 12 |
from doctr.models import ocr_predictor
|
| 13 |
-
|
|
|
|
| 14 |
# util functions
|
| 15 |
|
| 16 |
def prepreocess(text):
|
|
@@ -64,8 +65,9 @@ predictor = ocr_predictor(pretrained=True)
|
|
| 64 |
|
| 65 |
def proceed(image):
|
| 66 |
print("image: ", image)
|
| 67 |
-
|
| 68 |
-
|
|
|
|
| 69 |
result = predictor(doc) # extract words
|
| 70 |
json_export = result.export() # convert to json
|
| 71 |
out_words = [[prepreocess(w['value']) for w in line['words'] ] for line in json_export['pages'][0]['blocks'][0]['lines']] # extracting words
|
|
|
|
| 10 |
import matplotlib.pyplot as plt
|
| 11 |
from doctr.io import DocumentFile
|
| 12 |
from doctr.models import ocr_predictor
|
| 13 |
+
from PIL import Image
|
| 14 |
+
import numpy as np
|
| 15 |
# util functions
|
| 16 |
|
| 17 |
def prepreocess(text):
|
|
|
|
| 65 |
|
| 66 |
def proceed(image):
|
| 67 |
print("image: ", image)
|
| 68 |
+
im = Image.fromarray(np.array(image))
|
| 69 |
+
im.save("temp.jpg")
|
| 70 |
+
doc = DocumentFile.from_images("temp.jpg") #reading input image
|
| 71 |
result = predictor(doc) # extract words
|
| 72 |
json_export = result.export() # convert to json
|
| 73 |
out_words = [[prepreocess(w['value']) for w in line['words'] ] for line in json_export['pages'][0]['blocks'][0]['lines']] # extracting words
|