Update ComfyUI/custom_nodes/ComfyUI-BrushNet/brushnet_nodes.py

ComfyUI/custom_nodes/ComfyUI-BrushNet/brushnet_nodes.py

@@ -487,7 +487,12 @@ class BlendInpaint:
         blurred = []
         for i in range(inpaint.shape[0]):
             height, width, _ = original[i].shape
-            x0, y0, cut_width, cut_height = origin[i]
+
+            if origin is not None:
+                x0, y0, cut_width, cut_height = origin[i]
+            else:
+                x0, y0 = 0, 0
+                cut_width, cut_height = width, height
 
             # Ensure cut dimensions don't exceed original image dimensions
             cut_width = min(cut_width, width - x0)
@@ -518,7 +523,7 @@ class BlendInpaint:
 
 
 
-def scale_mask_and_image(image, mask, width, height):
+def scale_mask_and_image(image, mask, width, height, side_margin):
     h0, w0 = mask.shape
     iy, ix = (mask == 1).nonzero(as_tuple=True)
     
@@ -541,7 +546,7 @@ def scale_mask_and_image(image, mask, width, height):
         new_mask_height = mask_height
         new_mask_width = mask_height * aspect_ratio
 
-    margin = 0.1
+    margin = side_margin/100.0
     cut_width = int(new_mask_width * (1 + 2 * margin))
     cut_height = int(new_mask_height * (1 + 2 * margin))
 
@@ -572,8 +577,7 @@ class CutForInpaint:
                     {    
                         "image": ("IMAGE",),
                         "mask": ("MASK",),
-                        "width": ("INT", {"default": 512, "min": 64, "max": 2048}),
-                        "height": ("INT", {"default": 512, "min": 64, "max": 2048}),
+                        "side_margin_percent": ("INT", {"default": 10, "min": 0, "max": 1000})
                      },
                 }
 
@@ -583,12 +587,12 @@ class CutForInpaint:
 
     FUNCTION = "cut_for_inpaint"
 
-    def cut_for_inpaint(self, image: torch.Tensor, mask: torch.Tensor, width: int, height: int):
+    def cut_for_inpaint(self, image: torch.Tensor, mask: torch.Tensor, side_margin_percent: int):
         ret = []
         msk = []
         org = []
         for i in range(image.shape[0]):
-            cut_image, cut_mask, (x0, y0, cut_width, cut_height) = scale_mask_and_image(image[i], mask[i], width, height)
+            cut_image, cut_mask, (x0, y0, cut_width, cut_height) = scale_mask_and_image(image[i], mask[i], 512, 512, side_margin_percent)
             ret.append(cut_image)
             msk.append(cut_mask)
             org.append(torch.IntTensor([x0, y0, cut_width, cut_height]))
@@ -723,31 +727,6 @@ def prepare_image(image, mask):
     return (masked_image, mask)
 
 
-# Get origin of the mask
-def cut_with_mask(mask, width, height):
-    iy, ix = (mask == 1).nonzero(as_tuple=True)
-    h0, w0 = mask.shape
-    
-    if iy.numel() == 0:
-        x_c, y_c = w0 / 2.0, h0 / 2.0
-        mask_width, mask_height = 0, 0
-    else:
-        x_min, x_max = ix.min().item(), ix.max().item()
-        y_min, y_max = iy.min().item(), iy.max().item()
-        x_c, y_c = (x_min + x_max) / 2.0, (y_min + y_max) / 2.0
-        mask_width, mask_height = x_max - x_min + 1, y_max - y_min + 1
-
-    cut_width = max(width, mask_width * 1.4)  # 140% of mask width
-    cut_height = max(height, mask_height * 1.4)  # 140% of mask height
-
-    cut_width = min(cut_width, w0)
-    cut_height = min(cut_height, h0)
-
-    x0 = max(0, min(w0 - cut_width, x_c - cut_width / 2))
-    y0 = max(0, min(h0 - cut_height, y_c - cut_height / 2))
-
-    return (int(x0), int(y0), int(cut_width), int(cut_height))
-
 
 # Prepare conditioning_latents
 @torch.inference_mode()