top of page

Comfyui automation - AI-Powered Pipeline for Spine 2D Assets

An efficient automated ComfyUI workflow to generate and prepare modular 2D assets optimized for Spine skeletal animation. The pipeline integrates AI image generation, precise background removal and smart segmentation to output clean, production-ready PNG assets.

Allowing rapid iteration on new character designs or style variations.

Creative process

step 1. Background Removal & Image Splitting set up

Background Removal & Image Splitting Node built in ComfyUI

(Image background can be removed automatically and split into different parts as PNG)

Dramatically reduces manual labor and enhancing workflow.
Extension used:

-comfyui-inspyrenet-rembg

-comfyui-impack-pack

-mqsquerade-nodes-comfyui

step 2: API integration (Nano banana pro)

Used Nano Banana custom nodes in ComfyUI, linked with API key,

to generate high-quality 2D character and object illustrations. 

(Using pose reference image as "image_2" above for better AI integration)

Custom node used:

-ru4ls/ComfyUI_Nano_Banana

step 3: Combining 2 workflow

The output image from nano banana is then link up to the background removal

and image splitting system.

step 4: output

The output images are separated accordingly and packed into

specific folder, allowing fast integration with Spine2D.

bottom of page