Load Image Based on Number (Mikey)

Documentation

• Class name: Load Image Based on Number
• Category: Mikey/Image
• Output node: False

The node loads an image from a specified directory based on a given index (seed). It is designed to handle large indices by wrapping around the list of image files in the directory, ensuring a valid image is always returned regardless of the seed value.

Input types

Required

• image_directory
  • Specifies the directory from which to load the image. It is crucial for locating and accessing the desired images.
  • Comfy dtype: STRING
  • Python dtype: str
• seed
  • Determines the specific image to load by serving as an index into the sorted list of image files within the directory. The seed ensures a deterministic selection of images.
  • Comfy dtype: INT
  • Python dtype: int

Output types

• image
  • Comfy dtype: IMAGE
  • The loaded image, processed and converted into a tensor format suitable for further manipulation or analysis.
  • Python dtype: torch.Tensor
• filename
  • Comfy dtype: STRING
  • The name of the file from which the image was loaded, providing a reference to the original image file.
  • Python dtype: str

Usage tips

• Infra type: CPU
• Common nodes: unknown

Source code

class LoadImgFromDirectoryBasedOnIndex:
    # given a directory of images, and the seed number
    # return the image which is the index of the list of files in the directory
    # use mod to wrap around the list of files because the seed can be a huge number
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {"image_directory": ("STRING", {"multiline": False, "placeholder": "Image Directory"}),
                             "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff})}}

    RETURN_TYPES = ('IMAGE', 'STRING')
    RETURN_NAMES = ('image', 'filename')
    FUNCTION = 'load'
    CATEGORY = 'Mikey/Image'

    def load(self, image_directory, seed):
        if not os.path.exists(image_directory):
            raise Exception(f"Image directory {image_directory} does not exist")

        files = [os.path.join(image_directory, f)
                 for f in os.listdir(image_directory)
                 if os.path.isfile(os.path.join(image_directory, f)) and f.endswith((".png", ".jpg", ".jpeg", ".webp", ".bmp", ".gif"))]
        # sort files by name
        files.sort()
        # wrap around the list of files
        offset = seed % len(files)
        filename = files[offset].split('/')[-1]
        img = Image.open(files[offset])
        img = pil2tensor(img)
        return (img, filename)