Image Apply Channel

Documentation

• Class name: ImageApplyChannel
• Category: Art Venture/Utils
• Output node: False

The ImageApplyChannel node is designed to modify specific color channels (Red, Green, Blue, Alpha) of a collection of images by applying given channel data to them. This functionality allows for precise control over the color composition and transparency of images, enabling customized image processing and manipulation.

Input types

Required

• images
  • The collection of images to be modified. This parameter is crucial as it determines the base images that will undergo channel modification.
  • Comfy dtype: IMAGE
  • Python dtype: torch.Tensor
• channel_data
  • The data to be applied to the specified channel of each image. This parameter directly influences the outcome by altering the specified channel's values across all images.
  • Comfy dtype: MASK
  • Python dtype: torch.Tensor
• channel
  • Specifies which color channel (Red, Green, Blue, Alpha) to modify in the images. This choice dictates the aspect of the images that will be altered.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str

Output types

• image
  • Comfy dtype: IMAGE
  • The modified images with the specified channel data applied. This output reflects the changes made to the original images' specified channels.
  • Python dtype: torch.Tensor

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class UtilImageApplyChannel:
    @classmethod
    def INPUT_TYPES(s):
        return {
            "required": {
                "images": ("IMAGE",),
                "channel_data": ("MASK",),
                "channel": (["R", "G", "B", "A"],),
            }
        }

    RETURN_TYPES = ("IMAGE",)
    CATEGORY = "Art Venture/Utils"
    FUNCTION = "image_apply_channel"

    def image_apply_channel(self, images: torch.Tensor, channel_data: torch.Tensor, channel):
        merged_images = []

        for image in images:
            image = image.cpu().clone()

            if channel == "A":
                if image.shape[2] < 4:
                    image = torch.cat([image, torch.ones((image.shape[0], image.shape[1], 1))], dim=2)

                image[:, :, 3] = channel_data
            elif channel == "R":
                image[:, :, 0] = channel_data
            elif channel == "G":
                image[:, :, 1] = channel_data
            else:
                image[:, :, 2] = channel_data

            merged_images.append(image)

        return (torch.stack(merged_images),)