Update app.py

app.py

@@ -23,6 +23,7 @@ logger = logging.getLogger(__name__)
 
 # Download and cache the font file
 def get_font():
+    """Download font for annotation if not available."""
     try:
         font_path = Path("Roboto-Regular.ttf")
         if not font_path.exists():
@@ -34,33 +35,44 @@ def get_font():
         logger.error(f"Error in get_font: {str(e)}")
         return None
 
-# Initialize EasyOCR Reader for French
+# Initialize EasyOCR Reader for French & English
 try:
     reader = easyocr.Reader(['fr', 'en'], gpu=False)
 except Exception as e:
     logger.error(f"Error initializing EasyOCR: {str(e)}")
 
 def ocr_extract_text_and_tables(image):
+    """Extract text and tables from an image."""
     try:
         if image is None:
             return "No image provided", "No image provided", None
 
-        # Convert to RGB if needed
+        # Ensure image is in the correct format
+        if isinstance(image, Image.Image):
+            image = np.array(image, dtype=np.uint8)  # Convert PIL to numpy
+
+        # If image has an alpha channel (RGBA), convert to RGB
         if len(image.shape) == 3 and image.shape[2] == 4:
             image = cv2.cvtColor(image, cv2.COLOR_RGBA2RGB)
-        
+        elif len(image.shape) == 2:  # If grayscale, convert to BGR
+            image = cv2.cvtColor(image, cv2.COLOR_GRAY2BGR)
+
         # Convert to grayscale for better OCR
         gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
 
         # Apply adaptive thresholding
         processed = cv2.adaptiveThreshold(gray, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 11, 2)
         
-        # 1. Extract general text using EasyOCR
+        # 1️⃣ Extract General Text using EasyOCR
         results = reader.readtext(processed)
         detected_text = [f"{text} (Confidence: {confidence:.2f})" for _, text, confidence in results]
 
-        # 2. Use img2table for structured table extraction
-        img = Img2TableImage(image)
+        # 2️⃣ Save image to a temporary file for `img2table`
+        temp_image_path = "temp_table_image.jpg"
+        cv2.imwrite(temp_image_path, image)
+
+        # 3️⃣ Use img2table for structured table extraction
+        img = Img2TableImage(temp_image_path)  # Use file path instead of np.ndarray
         ocr = TesseractOCR(lang="fra")
         tables = img.extract_tables(ocr=ocr)
 
@@ -71,7 +83,7 @@ def ocr_extract_text_and_tables(image):
         for i, df in enumerate(table_data):
             df.to_csv(f"extracted_table_{i+1}.csv", index=False)
 
-        # Annotate image with bounding boxes around detected text
+        # 4️⃣ Annotate Image with Bounding Boxes for Detected Text
         pil_image = Image.fromarray(image)
         draw = ImageDraw.Draw(pil_image)
 
@@ -91,10 +103,10 @@ def ocr_extract_text_and_tables(image):
     except Exception as e:
         return f"Error: {str(e)}", "Processing failed", None
 
-# Create Gradio interface
+# Create Gradio Interface
 iface = gr.Interface(
     fn=ocr_extract_text_and_tables,
-    inputs=gr.Image(type="numpy", label="Upload Image"),
+    inputs=gr.Image(type="pil", label="Upload Image"),  # Ensures PIL image input
     outputs=[
         gr.Textbox(label="Extracted Text (French)"),
         gr.Textbox(label="Extracted Tables"),