Blur Image (Fast)

Documentation

• Class name: BlurImageFast
• Category: image/filters
• Output node: False

The BlurImageFast node provides a fast and efficient way to apply Gaussian blur to images. It is designed to blur images by specifying the radius of the blur in both the x and y directions, allowing for customizable blur effects.

Input types

Required

• images
  • The 'images' parameter represents the images to be blurred. It is crucial for defining the input images on which the Gaussian blur effect will be applied.
  • Comfy dtype: IMAGE
  • Python dtype: torch.Tensor
• radius_x
  • The 'radius_x' parameter specifies the horizontal radius of the Gaussian blur. It determines the extent of blurring along the x-axis of the images.
  • Comfy dtype: INT
  • Python dtype: int
• radius_y
  • The 'radius_y' parameter specifies the vertical radius of the Gaussian blur. It determines the extent of blurring along the y-axis of the images.
  • Comfy dtype: INT
  • Python dtype: int

Output types

• image
  • Comfy dtype: IMAGE
  • The output is a blurred version of the input images, achieved through Gaussian blurring as specified by the radius_x and radius_y parameters.
  • Python dtype: torch.Tensor

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class BlurImageFast:
    def __init__(self):
        pass

    @classmethod
    def INPUT_TYPES(s):
        return {
            "required": {
                "images": ("IMAGE",),
                "radius_x": ("INT", {
                    "default": 1,
                    "min": 0,
                    "max": 1023,
                    "step": 1
                }),
                "radius_y": ("INT", {
                    "default": 1,
                    "min": 0,
                    "max": 1023,
                    "step": 1
                }),
            },
        }

    RETURN_TYPES = ("IMAGE",)
    FUNCTION = "blur_image"

    CATEGORY = "image/filters"

    def blur_image(self, images, radius_x, radius_y):

        if radius_x + radius_y == 0:
            return (images,)

        dx = radius_x * 2 + 1
        dy = radius_y * 2 + 1

        dup = copy.deepcopy(images.cpu().numpy())

        for index, image in enumerate(dup):
            dup[index] = cv2.GaussianBlur(image, (dx, dy), 0)

        return (torch.from_numpy(dup),)