Extensions/ComfyUI-Mosaic-Mask
ComfyUI Extension

ComfyUI-Mosaic-Mask

ComfyUI-Mosaic-Mask is an automatic tool designed to detect and mask mosaic areas in input images.

By okgo4·Created 2 years ago·Updated 6 days ago· 10
okgo4/ComfyUI-Mosaic-Mask
Nodes1
On cloudLocal install
CategoryMosaic Masking
Stars10
Updated6 days ago

Nodes (1)

MosaicMask
Mosaic Masking
Readme

ComfyUI-Mosaic-Mask

ComfyUI-Mosaic-Mask detects regular mosaic grids in images and returns the detected areas as a ComfyUI mask.

Example

Features

  • Detects mosaic-like grid patterns with OpenCV template matching.
  • Supports image batches.
  • Returns a standard 0.0 to 1.0 ComfyUI MASK.
  • Keeps the largest disconnected regions and optionally expands them.

Installation

cd ComfyUI/custom_nodes
git clone https://github.com/okgo4/ComfyUI-Mosaic-Mask.git
python -m pip install -r ComfyUI-Mosaic-Mask/requirements.txt

Restart ComfyUI after installation.

Usage

Load example.json or add MosaicMask from the Mosaic Masking category.

  • top_n: maximum number of disconnected regions to keep, ordered by area.
  • kernel_size: dilation kernel size. Use 0 or 1 for no expansion.
  • threshold: template matching threshold. Higher values reduce false positives but may miss mosaics.
  • min_grid_size / max_grid_size: bundled template range to search. The default is 10 through 20; lower the minimum to detect finer grids.

The node includes grid templates from size 5 through 20. It works best with axis-aligned, regular square mosaics. Rotated grids, perspective distortion, unsupported scales, and naturally repetitive textures may produce missed detections or false positives.

A smoothing node may still be added for softer mask edges, but it is no longer required to normalize the output.

Thanks

Special thanks to the mosaic_detector project.