Update app.py

app.py

@@ -976,7 +976,7 @@ def predict_with_paper(image, paper_size, offset, offset_unit, finger_clearance=
         # Mask paper area in input image first
         masked_input_image = mask_paper_area_in_image(image, paper_contour)
         
-        # NEW: Use YOLOWorld to detect object bounding box
+        # Use YOLOWorld to detect object bounding box
         yolo_world = get_yolo_world()
         if yolo_world is None:
             logger.warning("YOLOWorld model not available, proceeding with full image")
@@ -1005,13 +1005,19 @@ def predict_with_paper(image, paper_size, offset, offset_unit, finger_clearance=
                 x_max = min(masked_input_image.shape[1], x_max + margin)
                 y_max = min(masked_input_image.shape[0], y_max + margin)
                 
-                # Crop the masked image
-                cropped_image = masked_input_image[y_min:y_max, x_min:x_max]
-                crop_offset = (x_min, y_min)  # Store offset for mask realignment
-                logger.info(f"Cropped to box: ({x_min}, {y_min}, {x_max}, {y_max})")
-                
-                # Debug: Save cropped image
-                cv2.imwrite("./debug/cropped_image.jpg", cropped_image)
+                # Validate crop region
+                if x_max <= x_min or y_max <= y_min:
+                    logger.warning("Invalid crop region, proceeding with full image")
+                    cropped_image = masked_input_image
+                    crop_offset = (0, 0)
+                else:
+                    # Crop the masked image
+                    cropped_image = masked_input_image[y_min:y_max, x_min:x_max]
+                    crop_offset = (x_min, y_min)  # Store offset for mask realignment
+                    logger.info(f"Cropped to box: ({x_min}, {y_min}, {x_max}, {y_max})")
+                    
+                    # Debug: Save cropped image
+                    cv2.imwrite("./debug/cropped_image.jpg", cropped_image)
         
         # Remove background from cropped image
         orig_size = image.shape[:2]
@@ -1020,7 +1026,10 @@ def predict_with_paper(image, paper_size, offset, offset_unit, finger_clearance=
         
         # Resize mask to match cropped region and place back in original image space
         full_mask = np.zeros((orig_size[0], orig_size[1]), dtype=np.uint8)
-        resized_mask = cv2.resize(objects_mask, (cropped_image.shape[1], cropped_image.shape[0]))
+        if cropped_image.shape[:2] != processed_size:
+            resized_mask = cv2.resize(objects_mask, (cropped_image.shape[1], cropped_image.shape[0]))
+        else:
+            resized_mask = objects_mask
         full_mask[y_min:y_min+resized_mask.shape[0], x_min:x_min+resized_mask.shape[1]] = resized_mask
         
         # Remove paper area from mask to focus only on objects
@@ -1045,33 +1054,6 @@ def predict_with_paper(image, paper_size, offset, offset_unit, finger_clearance=
     except Exception as e:
         raise gr.Error(f"Error in object detection: {str(e)}")
 
-    # Rest of the function remains unchanged...
-    # [Keep existing code for dilation, contour extraction, DXF generation, etc.]
-        objects_mask = remove_bg(image)
-        processed_size = objects_mask.shape[:2]
-        
-        # Resize mask to match original image
-        objects_mask = cv2.resize(objects_mask, (image.shape[1], image.shape[0]))
-        
-        # Remove paper area from mask to focus only on objects
-        objects_mask = exclude_paper_area(objects_mask, paper_contour)
-        
-        # Check if we actually have object pixels after paper exclusion
-        object_pixels = np.count_nonzero(objects_mask)
-        if object_pixels < 1000:  # Minimum threshold
-            raise NoObjectDetectedError("No significant object detected after excluding paper area")
-        
-        # Validate single object
-        validate_single_object(objects_mask, paper_contour)
-        
-    except (MultipleObjectsError, NoObjectDetectedError) as e:
-        return (
-            None, None, None, None,
-            f"Error: {str(e)}"
-        )
-    except Exception as e:
-        raise gr.Error(f"Error in object detection: {str(e)}")
-
     # Apply edge rounding if specified
     rounded_mask = objects_mask.copy()
     
@@ -1097,8 +1079,8 @@ def predict_with_paper(image, paper_size, offset, offset_unit, finger_clearance=
         # Generate DXF - scaling_factor should be in mm/px for proper DXF units
         dxf, finger_polygons, original_polygons = save_dxf_spline(
             contours,
-            scaling_factor,  # This should be mm/px
-            processed_size[0],
+            scaling_factor,  # This should mm/px
+            orig_size[0],  # Use original image height instead of processed_size
             finger_clearance=(finger_clearance == "On")
         )
     except FingerCutOverlapError as e:
@@ -1112,7 +1094,7 @@ def predict_with_paper(image, paper_size, offset, offset_unit, finger_clearance=
         for poly in finger_polygons:
             try:
                 coords = np.array([
-                    (int(x / scaling_factor), int(processed_size[0] - y / scaling_factor))
+                    (int(x / scaling_factor), int(orig_size[0] - y / scaling_factor))
                     for x, y in poly.exterior.coords
                 ], np.int32).reshape((-1, 1, 2))