Update app.py

app.py

@@ -1,11 +1,19 @@
 import gradio as gr
 from transformers import pipeline
 import re
+from PIL import ImageFilter
 
 # Load the OCR pipeline from Hugging Face
-ocr_pipeline = pipeline("image-to-text", model="microsoft/trocr-base-stage1")
+ocr_pipeline = pipeline("image-to-text", model="microsoft/trocr-large-stage1")  # Use a different model
+
+def preprocess_image(image):
+    image = image.convert('L')  # Convert to grayscale
+    image = image.filter(ImageFilter.SHARPEN)  # Apply some filtering
+    return image
 
 def perform_ocr(image):
+    # Preprocess the image before OCR
+    image = preprocess_image(image)
     text = ocr_pipeline(image)[0]['generated_text']
     return text
 
@@ -45,4 +53,4 @@ def web_app():
     interface.launch()
 
 if __name__ == "__main__":
-    web_app()
+    web_app()