Upload app.py

app.py

@@ -7,6 +7,17 @@ import numpy as np
 import gradio as gr
 from paddleocr import PaddleOCR
 
+from PIL import Image
+
+def is_valid_image(path):
+    try:
+        img = Image.open(path)
+        img.verify()
+        return True
+    except:
+        return False
+
+
 ocr = PaddleOCR(use_angle_cls=True, lang='en', det_model_dir='models/det', rec_model_dir='models/rec', cls_model_dir='models/cls')
 
 def classify_background_color(avg_color, white_thresh=230, black_thresh=50, yellow_thresh=100):
@@ -41,13 +52,25 @@ def sample_border_color(image, box, padding=2):
 
 def detect_text_boxes(image, max_dim=1280):
     try:
+        # Check if image is valid
+        if image is None or not hasattr(image, 'shape'):
+            print("Invalid image. Skipping...")
+            return []
+
         # Resize large images to reduce memory load
         height, width = image.shape[:2]
         if max(height, width) > max_dim:
             scale = max_dim / float(max(height, width))
             image = cv2.resize(image, (int(width * scale), int(height * scale)))
 
-        results = PaddleOCR.ocr(image, cls=True)
+        # Ensure image is in RGB
+        if image.shape[2] == 1:
+            image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
+        elif image.shape[2] == 3:
+            image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
+
+        # Call PaddleOCR correctly
+        results = ocr.ocr(image, cls=True)
 
         if results is None or not results[0]:
             print("No OCR results found or OCR returned None.")
@@ -69,6 +92,7 @@ def detect_text_boxes(image, max_dim=1280):
         return []
 
 
+
 def remove_text_dynamic_fill(img_path, output_path):
     image = cv2.imread(img_path)
     if image is None: