Image Remove Background

Image Remove Background (rembg)

Remove image backgrounds with AI matting via rembg. Output is RGBA PNG with transparent background — ready for Godot sprites and UI.

Rules

When this skill applies, read and follow [skill-dependency-manager](../../rules/skill-dependency-manager.md) — run scripts as documented, install missing tools into .dependency/.

• Run remove_background.py through the rembg manifest entry (.dependency/rembg/.venv/). Never use host python, py, python3, or any interpreter outside .dependency/.
• Do not hand-write rembg i / rembg p — use the bundled script.
• populated: false for rembg is not a reason to skip. Install first, set populated: true, retry the same command.
• Pass the input path as-is (chat attachment path, Downloads/foo.png, project folder, etc.). Output goes to /transparent/ by default — no path rewriting.

Setup (first run)

From project root:

.dependency/python/python -m venv .dependency/rembg/.venv
.dependency/rembg/.venv/Scripts/python -m pip install "rembg[cpu]"

GPU (CUDA) — faster batch processing:

.dependency/rembg/.venv/Scripts/python -m pip install "rembg[gpu]"

Register in .dependency/manifest.json:

"rembg": {
  "populated": true,
  "bin": ".dependency/rembg/.venv/Scripts/python.exe"
}

Use bin/python on Unix. Model weights download on first run (~hundreds of MB).

Quick Start

Default: create a transparent/ folder under the input path and write outputs there (never overwrites sources):

# Single file → image/sprites/transparent/hero.png
.dependency/rembg/.venv/Scripts/python .cursor/skills/image-remove-background/scripts/remove_background.py image/sprites/hero.png

# Directory batch → image/sprites/hero/transparent/.png
.dependency/rembg/.venv/Scripts/python .cursor/skills/image-remove-background/scripts/remove_background.py image/sprites/hero -r
# e.g. image/sprites/hero/sub/foo.png → image/sprites/hero/transparent/sub/foo.png

Custom output directory:

.dependency/rembg/.venv/Scripts/python .cursor/skills/image-remove-background/scripts/remove_background.py image/sprites/hero \
  --output-dir image/sprites/hero_cutout

Model selection

| Model | Use case | |-------|----------| | u2net (default) | General objects, icons, props | | u2netp | Faster / lighter; smaller assets | | isnet-general-use | Higher quality general matting | | birefnet-general | Best general quality (slower) | | birefnet-portrait | Characters / portraits | | u2net_human_seg | Human figures only |

.dependency/rembg/.venv/Scripts/python .cursor/skills/image-remove-background/scripts/remove_background.py image/characters \
  --model birefnet-portrait --output-dir image/characters_cutout

Edge quality (alpha matting)

For hair, fur, or soft edges, enable alpha matting:

.dependency/rembg/.venv/Scripts/python .cursor/skills/image-remove-background/scripts/remove_background.py image/portrait.png \
  --alpha-matting --output-dir image/portrait_cutout

Defaults

| Option | Default | Notes | |--------|---------|-------| | Output | /transparent/ | Folder is auto-created under the file or directory you pass; use --output-dir for a custom path | | --model | u2net | See table above | | --pattern | *.png | Also matches .jpg, .jpeg, .webp | | --alpha-matting | off | Enable for fine edge detail | | --crop | off | Trim transparent borders after matting | | Overwrite | off | Pass --overwrite to replace existing outputs |

Agent workflow

1. Paths — Pass whatever path the user gives or the chat ` path directly. Output lands in transparent/` next to that input.
2. Trial first — run on 1 image, inspect the transparent/ or --output-dir result before batch.
3. Pick model — u2net for generic assets; birefnet-portrait for characters; birefnet-general when quality matters.
4. Soft edges — try --alpha-matting if halos or jagged hair/fur appear.
5. Sprite sheets — skip *_sheet.png by default; process individual frames unless the user asks otherwise.
6. Already transparent — script still runs; rembg re-mats from visible RGB. Warn user if source already has alpha.
7. Revert — delete output folder or git restore if needed; sources are never modified.

Troubleshooting

| Issue | Fix | |-------|-----| | rembg missing | Follow Setup; update manifest | | Very slow | Install rembg[gpu]; try --model u2netp | | Jagged edges | --alpha-matting | | Wrong subject removed | Switch model; try birefnet-general | | Leftover background color | Re-run with --alpha-matting; check source contrast | | OOM on large images | Script auto-downscales inputs above 4096 px longest side |

Related

• Flat white/green/magenta AI backgrounds: [image-remove-white-background](../image-remove-white-background/SKILL.md) (prefer over rembg)
• rembg docs: https://github.com/danielgatis/rembg

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: godot-fun
• Source: godot-fun/godot-framework
• License: MIT

Install and usage instructions live in the source repository linked above.