LayerStyle

2167
By chflame
Updated 15 days ago
View on GitHub →See Common Issues →

A set of nodes for ComfyUI that can composite layer and mask to achieve Photoshop like functionality.

Available Nodes

Repository Overview

LayerMask: RmBgUltra V2

LayerMask: RmBgUltra V2 Node Documentation

Overview

The LayerMask: RmBgUltra V2 node in the ComfyUI framework is designed to remove backgrounds from images and generate corresponding masks using advanced image processing techniques. This node is particularly useful for workflows that require image segmentation or background removal for further processing or image composition tasks.

Functionality

This node processes each input image to create a version of the image with a removed background, alongside a mask that corresponds to the original image. It achieves this through a combination of image processing methods, which can be tailored to suit specific requirements in terms of detail and processing method.

Inputs

The node accepts the following input parameters:

  • Image: The source image(s) from which the background will be removed.
  • Detail Method: Allows you to select the method used for detail processing. Options include:
    • VITMatte
    • VITMatte(local)
    • PyMatting
    • GuidedFilter
  • Detail Erode: Controls the amount of erosion applied to the detail mask. Accepts integer values (1-255), with a default of 6.
  • Detail Dilate: Controls the amount of dilation applied to the detail mask. Accepts integer values (1-255), with a default of 6.
  • Black Point: Defines the low-end threshold for the detail mask's histogram. Accepts float values (0.01-0.98) with a slider for adjustment, defaulting to 0.01.
  • White Point: Defines the high-end threshold for the detail mask's histogram. Accepts float values (0.02-0.99) with a slider for adjustment, defaulting to 0.99.
  • Process Detail: Boolean to specify whether to process detail or not, defaulting to true.
  • Device: Determines the processing hardware (cuda or cpu).
  • Max Megapixels: Limits the maximum size of the processed image in megapixels, accepting float values between 1 and 999, with a default value of 2.0.

Outputs

The node produces two main outputs:

  • Image: An image where the background has been removed, typically in RGBA format (with alpha representing transparency).
  • Mask: A mask image corresponding to the removed background. This mask can be used for further operations like blending or compositing in other environments or workflows.

Usage in ComfyUI Workflows

In a ComfyUI workflow, the LayerMask: RmBgUltra V2 node can be integrated wherever background removal is required. This could be useful in scenarios such as:

  • Image Compositing: Combine foreground elements with different backgrounds.
  • Image Analysis: Focus on specific regions of interest by filtering out unnecessary background details.
  • Enhancement and Retouching: Remove distracting background elements to improve the visual quality of an image.

Users should select the appropriate detail method based on the specific requirements of their application, such as processing speed or the level of detail needed.

Special Features and Considerations

  • Processing Method Flexibility: The node supports multiple detail processing methods offering flexibility depending on the desired outcome and available processing power.
  • Customization Options: Users can configure detail erosion and dilation, black and white points, and decide whether to process the detail, allowing for tailored image processing.
  • Device Selection: Supports both CPU and CUDA devices, making it adaptable to different hardware setups.
  • Efficiency: The option to limit the maximum megapixels ensures efficient processing, especially with larger data sets.

This node provides a comprehensive solution for image background removal and is a valuable tool in any image processing or enhancement workflow within the ComfyUI environment.