LayerFilter: GaussianBlur

Documentation

• Class name: LayerFilter: GaussianBlur
• Category: 😺dzNodes/LayerFilter
• Output node: False

The GaussianBlur node applies a Gaussian blur filter to images, effectively smoothing them by reducing high-frequency noise and details. This node is part of the LayerFilter category, focusing on enhancing or modifying the visual appearance of images through blurring techniques.

Input types

Required

• image
  • The 'image' parameter represents the input image(s) to be blurred. It is crucial for defining the visual content that will undergo the Gaussian blur transformation.
  • Comfy dtype: IMAGE
  • Python dtype: torch.Tensor
• blur
  • The 'blur' parameter specifies the intensity of the Gaussian blur effect. A higher value results in a more pronounced blurring effect, directly influencing the smoothness of the output image.
  • Comfy dtype: INT
  • Python dtype: int

Optional

Output types

• image
  • Comfy dtype: IMAGE
  • The output 'image' parameter is the result of applying the Gaussian blur filter to the input image(s), showcasing the transformed visual content with reduced noise and details.
  • Python dtype: torch.Tensor

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class GaussianBlur:

    def __init__(self):
        pass

    @classmethod
    def INPUT_TYPES(self):

        return {
            "required": {
                "image": ("IMAGE", ),  #
                "blur": ("INT", {"default": 20, "min": 1, "max": 999, "step": 1}),  # 模糊
            },
            "optional": {
            }
        }

    RETURN_TYPES = ("IMAGE",)
    RETURN_NAMES = ("image",)
    FUNCTION = 'gaussian_blur'
    CATEGORY = '😺dzNodes/LayerFilter'

    def gaussian_blur(self, image, blur):

        ret_images = []

        for i in image:
            _canvas = tensor2pil(torch.unsqueeze(i, 0)).convert('RGB')

            ret_images.append(pil2tensor(gaussian_blur(_canvas, blur)))

        log(f"{NODE_NAME} Processed {len(ret_images)} image(s).", message_type='finish')
        return (torch.cat(ret_images, dim=0),)