Update app.py

app.py

@@ -3,6 +3,7 @@ import torch
 import pytesseract
 from transformers import AutoTokenizer, AutoModel
 
+# Set Tesseract executable path
 pytesseract.pytesseract.tesseract_cmd = r'/opt/homebrew/bin/tesseract'
 
 # Load the tokenizer and model
@@ -10,19 +11,27 @@ tokenizer_eng = AutoTokenizer.from_pretrained('ucaslcl/GOT-OCR2_0', trust_remote
 model_eng = AutoModel.from_pretrained('ucaslcl/GOT-OCR2_0', trust_remote_code=True).eval()
 
 def perform_ocr(image, language):
-    # Perform OCR for English
+    # Convert the Gradio image input to the format suitable for pytesseract
     img_cv = image  # Assuming image is already in the correct format
-    res_eng = model_eng.chat(tokenizer_eng, img_cv, ocr_type='ocr')
 
-    return res_eng  # Return results for English
+    if language == "English":
+        # Perform OCR using the model for English
+        res_eng = model_eng.chat(tokenizer_eng, img_cv, ocr_type='ocr')
+        return res_eng  # Return results for English
+    elif language == "Hindi":
+        # Perform OCR using pytesseract for Hindi
+        res_hin = pytesseract.image_to_string(img_cv, lang='hin', config='--psm 6')
+        return res_hin  # Return results for Hindi
+    else:
+        return "Unsupported language selected."
 
 def ocr_and_search(image, language):
     # Call the perform_ocr function
-    english_text = perform_ocr(image, language)
+    extracted_text = perform_ocr(image, language)
     # You may also want to implement any searching functionality here
     # ...
     
-    return english_text  # Return the OCR result for English
+    return extracted_text  # Return the OCR result for the selected language
 
 # Create Gradio interface
 iface = gr.Interface(