Create app.py

app.py

@@ -0,0 +1,93 @@
+from pdf2image import convert_from_path
+import base64
+from io import BytesIO
+import pandas as pd
+import PIL
+import numpy as np
+from PIL import Image
+from PIL import ImageDraw
+import gradio as gr
+import torch
+import easyocr
+
+torch.hub.download_url_to_file('https://github.com/JaidedAI/EasyOCR/raw/master/examples/english.png', 'english.png')
+torch.hub.download_url_to_file('https://github.com/JaidedAI/EasyOCR/raw/master/examples/japanese.jpg', 'japanese.jpg')
+torch.hub.download_url_to_file('https://i.imgur.com/mwQFd7G.jpeg', 'Hindi.jpeg')
+
+def draw_boxes(image, bounds, color='yellow', width=2):
+    draw = ImageDraw.Draw(image)
+    for bound in bounds:
+        p0, p1, p2, p3 = bound[0]
+        draw.line([*p0, *p1, *p2, *p3, *p0], fill=color, width=width)
+    return image
+
+
+def inference(img, lang):
+    global img1,bounds
+    img1= img  
+    reader = easyocr.Reader(lang)
+    bounds = reader.readtext(img)
+    try:
+      #bounds = reader.readtext(img)
+      im = PIL.Image.open(img)
+    except:
+      im = PIL.Image.fromarray(img)
+    draw_boxes(im, bounds)
+    cdf = pd.DataFrame([x[1:] for x in bounds],columns=['text','confidence'])
+    cdf['confidence'] = cdf['confidence'].apply(lambda x:"{0:.1%}".format(x))
+    return [im, cdf]
+
+def read_from_file(file_path,lang):
+    if '.pdf' in str(file_path):
+      print('pdf')
+      # Convert each page of the PDF into an image
+      pages = convert_from_path(file_path)
+
+      # Convert each page image to base64 and append to the list
+      data = []
+      imgs = []
+      n = 0
+      for i, page in enumerate(pages):
+          if n>=5:
+            break
+          # Save the page image to a byte buffer
+          img,dfx = inference(np.array(page)[:, :, ::-1],lang)
+          imgs.append(img)
+          dfx['page'] = n
+          data.append(dfx)
+          n+=1
+      data = pd.concat(data).reset_index(drop=True)
+      data
+      return [imgs,data]
+    else:
+      print('image')
+      img2 = np.array(PIL.Image.open(file_path))
+      a,b = inference(img2,lang)
+      return [[a],b] 
+
+
+title = 'Gradio OCR Demo'
+description = 'Gradio OCR demo supports 80+ languages. To use it, simply upload your image and choose a language from the dropdown menu, or click one of the examples to load them.'
+examples = [['english.png',['en']],['japanese.jpg',['ja', 'en']],['Hindi.jpeg',['hi', 'en']]]
+css = ".output_image, .input_image {height: 40rem !important; width: 100% !important;}"
+choices = ['abq', 'ady', 'af', 'ang', 'ar', 'as', 'ava', 'az', 'be', 'bg',
+       'bh', 'bho', 'bn', 'bs', 'ch_sim', 'ch_tra', 'che', 'cs', 'cy',
+       'da', 'dar', 'de', 'en', 'es', 'et', 'fa', 'fr', 'ga', 'gom', 'hi',
+       'hr', 'hu', 'id', 'inh', 'is', 'it', 'ja', 'kbd', 'kn', 'ko', 'ku',
+       'la', 'lbe', 'lez', 'lt', 'lv', 'mah', 'mai', 'mi', 'mn', 'mr',
+       'ms', 'mt', 'ne', 'new', 'nl', 'no', 'oc', 'pi', 'pl', 'pt', 'ro',
+       'ru', 'rs_cyrillic', 'rs_latin', 'sck', 'sk', 'sl', 'sq', 'sv',
+       'sw', 'ta', 'tab', 'te', 'th', 'tjk', 'tl', 'tr', 'ug', 'uk', 'ur',
+       'uz', 'vi']
+app = gr.Interface(
+    read_from_file,
+    inputs=[gr.File(type='filepath',label='input_file pdf or image'),#gr.Image(type='filepath', label='Input'),
+            gr.CheckboxGroup(choices, type="value", label='language')],
+    outputs=[gr.Gallery(columns=4), 'dataframe'],
+    title=title,
+    description=description,
+    examples=examples,
+    css=css
+    )
+
+app.launch(debug=True,enable_queue=True)