Files
2025-01-02 15:54:50 +09:00

98 lines
4.3 KiB
Python

import os
import torch
from PIL import Image
import torchvision.transforms.functional as tf
import logging
import numpy as np
import scipy.ndimage
import comfy.utils
MAX_RESOLUTION=16384
def composite(destination, source, x, y, mask = None, multiplier = 8, resize_source = False):
source = source.to(destination.device)
if resize_source:
source = torch.nn.functional.interpolate(source, size=(destination.shape[2], destination.shape[3]), mode="bilinear")
source = comfy.utils.repeat_to_batch_size(source, destination.shape[0])
x = max(-source.shape[3] * multiplier, min(x, destination.shape[3] * multiplier))
y = max(-source.shape[2] * multiplier, min(y, destination.shape[2] * multiplier))
left, top = (x // multiplier, y // multiplier)
right, bottom = (left + source.shape[3], top + source.shape[2],)
if mask is None:
mask = torch.ones_like(source)
else:
mask = mask.to(destination.device, copy=True)
mask = torch.nn.functional.interpolate(mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])), size=(source.shape[2], source.shape[3]), mode="bilinear")
mask = comfy.utils.repeat_to_batch_size(mask, source.shape[0])
# calculate the bounds of the source that will be overlapping the destination
# this prevents the source trying to overwrite latent pixels that are out of bounds
# of the destination
visible_width, visible_height = (destination.shape[3] - left + min(0, x), destination.shape[2] - top + min(0, y),)
mask = mask[:, :, :visible_height, :visible_width]
inverse_mask = torch.ones_like(mask) - mask
source_portion = mask * source[:, :, :visible_height, :visible_width]
destination_portion = inverse_mask * destination[:, :, top:bottom, left:right]
print('source_portion:', source_portion.shape)
print('destination_portion:', destination_portion.shape)
destination[:, :, top:bottom, left:right] = source_portion + destination_portion
return destination
class StickerMaskComposite:
@classmethod
def INPUT_TYPES(s):
return {
"required": {
"destination": ("IMAGE",),
"source": ("IMAGE",),
"x1": ("INT", {"default": 94, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"y1": ("INT", {"default": 700, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"x2": ("INT", {"default": 614, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"y2": ("INT", {"default": 700, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"x3": ("INT", {"default": 1134, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"y3": ("INT", {"default": 700, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"x4": ("INT", {"default": 94, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"y4": ("INT", {"default": 1420, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"x5": ("INT", {"default": 614, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"y5": ("INT", {"default": 1420, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"x6": ("INT", {"default": 1134, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"y6": ("INT", {"default": 1420, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
"resize_source": ("BOOLEAN", {"default": False}),
},
"optional": {
"mask": ("MASK",),
}
}
RETURN_TYPES = ("IMAGE",)
FUNCTION = "composite"
CATEGORY = "image"
def composite(self, destination, source, x1, y1, x2, y2, x3, y3, x4, y4, x5, y5, x6, y6, resize_source, mask = None):
# print(source.shape)
size_pack = [(x1, y1), (x2, y2), (x3, y3), (x4, y4), (x5, y5), (x6, y6)]
# print("dest", destination.shape)
for i in range(len(size_pack)):
destination = destination.clone().movedim(-1, 1)
# print("source", source[i].unsqueeze(0).movedim(-1, 1).shape)
destination = composite(destination, source[i].unsqueeze(0).movedim(-1, 1), size_pack[i][0], size_pack[i][1], mask[i].unsqueeze(0), 1, resize_source).movedim(1, -1)
return (destination,)
NODE_CLASS_MAPPINGS = {
"Sticker_Compositer": StickerMaskComposite,
}
NODE_DISPLAY_NAME_MAPPINGS = {
"Sticker_Compositer": "Sticker Compositer",
}