Update app.py

app.py

@@ -41,8 +41,12 @@ def vision_ai_api(image, label):
     }
 
 def predict(image):
-    image = preprocess_image(image)
-    
+    # Convert PIL image to NumPy array
+    if isinstance(image, Image.Image):
+        image = np.array(image)
+
+    image = preprocess_image(image)  # Apply preprocessing
+
     results = model(image, conf=0.80)
 
     detected_classes = set()
@@ -62,19 +66,22 @@ def predict(image):
             labels.append(f"{class_name} {conf:.2f}")
 
             # Ensure bounding boxes are within the image
-            height, width = image.shape[:2]
+            height, width = image.shape[:2]  # ✅ This now works
             x1, y1, x2, y2 = max(0, x1), max(0, y1), min(width, x2), min(height, y2)
 
             if x1 >= x2 or y1 >= y2:
                 print("Invalid bounding box, skipping.")
                 continue
 
-            cropped = image[y1:y2, x1:x2]
-            cropped_pil = Image.fromarray(cropped)
+            cropped = image[y1:y2, x1:x2]  # Crop the detected region
+            cropped_pil = Image.fromarray(cropped)  # Convert back to PIL
 
             # Call API
             api_response = vision_ai_api(cropped_pil, class_name)
-            cropped_images[class_name] = {"image": cropped_pil, "api_response": json.dumps(api_response, indent=4)}
+            cropped_images[class_name] = {
+                "image": cropped_pil,
+                "api_response": json.dumps(api_response, indent=4)
+            }
 
     if not cropped_images:
         return None, "No front detected", None, "No back detected", ["No valid detections"]
@@ -88,6 +95,7 @@ def predict(image):
     )
 
 
+
 # Gradio Interface
 iface = gr.Interface(
     fn=predict,