comfyui-load_image_enhanced-plugin

ComfyUI LoadImageEnhanced

This custom node extends ComfyUI's image loading functionality with filename output and original filename tracking.

Features

LoadImageEnhanced

• Filename Output: Returns the filename of the loaded image as a STRING output (the original LoadImage node has no filename output)
• Original Filename Tracking: Preserves the original filename even when the image is modified by tools like MaskEditor (toggleable)
• Sync MaskEditor Toggle: sync_maskeditor input controls whether original_filename follows MaskEditor writes
• Recursive Subfolder Listing: Lists images from all subfolders (the original LoadImage node only lists the top-level input directory); excludes clipspace, 3d, pasted, and any folder/file starting with .

Installation

ComfyUI Manager / Comfy Registry (recommended)

The node is published to the Comfy Registry as load-image-enhanced.

1. Open the ComfyUI Manager (or the Comfy Registry / extensions page).
2. Search for load-image-enhanced.
3. Click Install.
4. Restart ComfyUI.

Manual (git clone)

1. Clone the repository into your ComfyUI custom_nodes directory:

   git clone https://github.com/yusrmuttaqien/comfyui-load_image_enhanced-plugin.git custom_nodes/load-image-enhanced
   

2. Restart ComfyUI.
3. The new node will appear in the "image" category.

Usage

LoadImageEnhanced

• Input: Select an image file from the dropdown (includes images from subfolders)
• Inputs:
  • image — Image filename to load
  • original_filename — Text field that stores the original selected filename (editable by user)
  • sync_maskeditor — Toggle (default off): when off, original_filename ignores MaskEditor writes; when on, it follows every source including MaskEditor
• Outputs:
  • image: The loaded image tensor
  • mask: The image mask (if available)
  • filename: The filename of the loaded image

Requirements

• ComfyUI
• PIL (Pillow)
• PyTorch
• NumPy

Notes

Filename Tracking

The filename output uses the value of the original_filename widget, falling back to the loaded file's basename when the widget is empty. The widget is automatically updated when you select a new image from the dropdown. Whether it also follows MaskEditor writes is controlled by the sync_maskeditor toggle:

• sync_maskeditor off (default) — original_filename ignores MaskEditor (clipspace) writes:
  • Select characters/hero.png → filename output is hero.png
  • MaskEditor saves to clipspace-painted-xxx.png → the dropdown updates, but original_filename stays hero.png
  • The filename output continues to return hero.png
• sync_maskeditor on — original_filename follows every source, including MaskEditor:
  • MaskEditor saves to clipspace-painted-xxx.png → original_filename updates to clipspace-painted-xxx.png

Frontend Extension (Automatic)

The node includes a frontend extension that automatically syncs the original_filename widget when you change the dropdown. It's bundled with the custom node and loads automatically via WEB_DIRECTORY.

Without the extension, the original_filename field is still available but must be manually edited.

• The node recursively lists all images in the input folder and its subfolders
• Images inside clipspace, 3d, pasted, and any hidden (starting with .) folders are excluded from the dropdown listing
• Any file starting with . (e.g., .DS_Store) is also excluded from the listing
• However, if a downstream node (e.g., mask editor) passes an image path from an excluded folder, the node will still load it correctly

Based On

• ComfyUI LoadImage Node