Load Image (Inspire)

Documentation

• Class name: LoadImage __Inspire
• Category: InspirePack/image
• Output node: False

The LoadImage node in the Inspire pack is designed to load and preprocess images for further use in image processing workflows. It handles image loading from a specified directory, applying necessary transformations such as decoding, orientation correction, and normalization, and optionally generating a mask for images with transparency.

Input types

Required

• image
  • Specifies the image file to be loaded. This input is crucial for determining which image is processed and loaded into the workflow.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str
• image_data
  • Provides the image data in a base64-encoded string format. This data is decoded and processed to produce the image and its mask if applicable, serving as an alternative input method for image loading.
  • Comfy dtype: STRING
  • Python dtype: str

Output types

• image
  • Comfy dtype: IMAGE
  • The processed image, ready for use in subsequent image processing tasks. It is returned as a normalized tensor, reflecting the applied preprocessing steps.
  • Python dtype: torch.Tensor
• mask
  • Comfy dtype: MASK
  • An optional output providing a mask for the image, useful for images with transparency. It indicates areas of interest or exclusion in further processing steps.
  • Python dtype: torch.Tensor

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class LoadImageInspire:
    @classmethod
    def INPUT_TYPES(s):
        input_dir = folder_paths.get_input_directory()
        files = [f for f in os.listdir(input_dir) if os.path.isfile(os.path.join(input_dir, f))]
        return {"required": {
                                "image": (sorted(files) + ["#DATA"], {"image_upload": True}),
                                "image_data": ("STRING", {"multiline": False}),
                            }
                }

    CATEGORY = "InspirePack/image"

    RETURN_TYPES = ("IMAGE", "MASK")
    FUNCTION = "load_image"

    def load_image(self, image, image_data):
        image_data = base64.b64decode(image_data.split(",")[1])
        i = Image.open(BytesIO(image_data))
        i = ImageOps.exif_transpose(i)
        image = i.convert("RGB")
        image = np.array(image).astype(np.float32) / 255.0
        image = torch.from_numpy(image)[None,]
        if 'A' in i.getbands():
            mask = np.array(i.getchannel('A')).astype(np.float32) / 255.0
            mask = 1. - torch.from_numpy(mask)
        else:
            mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu")
        return (image, mask.unsqueeze(0))