Image Resize To Height

Documentation

• Class name: Image Resize To Height
• Category: Hakkun
• Output node: False

This node is designed to resize an image to a specified target height while maintaining its aspect ratio. It calculates the necessary scaling factor based on the target height and the original height of the image, then applies a Lanczos filter for high-quality resizing.

Input types

Required

• image
  • The image to be resized. It is crucial for determining the new dimensions of the image after resizing to the target height.
  • Comfy dtype: IMAGE
  • Python dtype: torch.Tensor
• target_height
  • The desired height of the image after resizing. This parameter directly influences the scaling factor used to adjust the image's height.
  • Comfy dtype: INT
  • Python dtype: int

Output types

• image
  • Comfy dtype: IMAGE
  • The resized image with the specified target height, maintaining the original aspect ratio.
  • Python dtype: torch.Tensor

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class ImageResizeToHeight:
    def __init__(self):
        pass

    @classmethod
    def INPUT_TYPES(cls):
        return {
            "required": {
                "image": (IMAGE_TYPE,),
                "target_height": (INT_TYPE, {"default": 1920, "min": 1, "step": 1}),
            }
        }

    RETURN_TYPES = (IMAGE_TYPE,)
    RETURN_NAMES = ("image",)
    FUNCTION = "calculate"

    CATEGORY = "Hakkun"

    def calculate(self, image, target_height):
        image_size = image.size()
        img_height = int(image_size[1])
        scale_by = target_height/img_height
        return upscale(image, 'lanczos', scale_by)