Update app.py

app.py

@@ -11,12 +11,13 @@ from huggingface_hub import snapshot_download
 import easyocr
 import re
 from typing import Tuple
+import numpy as np
 
 # Disable CUDA (Optional: as per your original code)
 torch.cuda.is_available = lambda: False
 
 # Set Streamlit page configuration
-st.set_page_config(page_title="GOT-OCR2_0 OCR App", layout="wide")
+st.set_page_config(page_title="Bilingual OCR App", layout="wide")
 
 @st.cache_resource
 def setup_got_model() -> Tuple[object, object]:
@@ -50,7 +51,7 @@ def perform_got_ocr(model, processor, image: Image.Image) -> str:
     generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
     return generated_text
 
-def perform_easyocr(ocr_reader, image: Image.Image) -> str:
+def perform_easyocr(ocr_reader, image: np.ndarray) -> str:
     results = ocr_reader.readtext(image, detail=0, paragraph=True)
     extracted_text = '\n'.join(results)
     return extracted_text
@@ -74,7 +75,7 @@ def main():
         ocr_reader = setup_easyocr()
         ocr_model = 'EasyOCR'
 
-    st.title("GOT-OCR2_0 OCR Application")
+    st.title("Bilingual OCR Application")
 
     # Sidebar for instructions
     st.sidebar.header("Instructions")
@@ -90,15 +91,23 @@ def main():
 
     if uploaded_file is not None:
         try:
-            image = Image.open(uploaded_file).convert('RGB')
-            st.image(image, caption="Uploaded Image", use_column_width=True)
+            # Read the file into bytes
+            image_bytes = uploaded_file.read()
+            
+            # Open the image with PIL
+            image_pil = Image.open(BytesIO(image_bytes)).convert('RGB')
+            
+            # Convert to numpy array for EasyOCR
+            image_np = np.array(image_pil)
+            
+            st.image(image_pil, caption="Uploaded Image", use_column_width=True)
 
             # Perform OCR
             with st.spinner(f"Performing OCR using {ocr_model}..."):
                 if ocr_model == 'GOT-OCR2_0' and model and processor:
-                    extracted_text = perform_got_ocr(model, processor, image)
+                    extracted_text = perform_got_ocr(model, processor, image_pil)
                 else:
-                    extracted_text = perform_easyocr(ocr_reader, image)
+                    extracted_text = perform_easyocr(ocr_reader, image_np)
 
             st.subheader("Extracted Text:")
             st.text_area("Text", extracted_text, height=200)
@@ -118,4 +127,4 @@ def main():
         st.info("Please upload an image file to get started.")
 
 if __name__ == "__main__":
-    main()
+    main()