De Una · Banco Pichincha
Escanea el QR con tu app bancaria para enviar tu aporte.
¡Muchas gracias!
Erase people, edit only part of a picture, reshape, curves and film looks — as easy as a phone app, but non-destructive and saved at your photo's original resolution. No downscaling, no re-compression.
The power lives under a simple, direct-on-image interface. Do a little or a lot.
Preview is fast, but Save always renders at your photo's native resolution — real JPEG quality or lossless PNG/TIFF. Your edits stack as layers; the original pixels are never thrown away.
Brush a region and adjust just that area. A feather control and edge-aware masking blend the change so it never looks cut out.
Paint over a distraction and it's gone. Small blemishes use fast classical inpainting; larger objects use LaMa AI, which invents a matching background and colour-matches the fill.
Brush roughly over someone and SAM segments them precisely, then LaMa fills the gap. It runs on a background thread, so the app never freezes.
A liquify push-brush to reshape a body or object — drag an edge inward to slim. Stored as a smooth warp field, so it stays non-destructive and full-resolution.
Exposure, brightness, contrast, highlights, shadows, saturation, vibrance, temperature, tint — plus a draggable master and per-channel R/G/B curve. All work globally or on a selection.
Camera profiles — Standard, Vivid, Neutral, Landscape, Portrait, Monochrome and more — alongside 20 Fujifilm-inspired film simulations.
Drop an object from another photo into the scene. It's Poisson-blended and colour-matched to its surroundings so the seam disappears.
Save an .iedit project to reopen your whole edit — layers, masks, curves — and keep tweaking later.
No presets to wrangle, no cloud round-trip. Just your photo at full quality.
JPEG, PNG, TIFF, BMP, WEBP. Unicode-safe paths on Linux, macOS, and Windows.
Move sliders and curves for the whole image, or brush a selection and edit just that region with a soft, blended border.
Erase distractions or a whole person, reshape and slim, or paste something in — seamlessly blended.
Export at native resolution, or save an .iedit project to keep editing later.
A Qt-free editing core (NumPy + OpenCV) drives both the live preview and the full-resolution export from the same code, so what you see is what you save.
| Adjustments | Exposure · Brightness · Contrast · Highlights · Shadows · Saturation · Vibrance · Temperature · Tint · Curves |
|---|---|
| Looks | 10 camera profiles + 20 film simulations (30 total) |
| Retouch | Local adjustments · Erase (classical + LaMa AI) · Remove person (SAM) · Reshape/liquify · Seamless paste |
| Output | JPEG (quality) · PNG · TIFF — always at the original resolution, saved atomically with the source date preserved |
| Runs | 100% local · CPU-only core · optional CPU AI models · no account, no telemetry, no cloud |
The core editor needs no models. The AI features (Erase / Remove person) are an optional install and degrade gracefully when absent — see the install guide.
Python 3.11+, Linux / macOS / Windows. No build step — clone and run.
# Clone and launch (first run sets up the venv + installs deps) git clone https://github.com/mnavas/image-editor.git cd image-editor ./launch.sh # …or by hand python3 -m venv .venv source .venv/bin/activate # Windows: .venv\Scripts\activate pip install -r requirements.txt python main.py # Optional AI (CPU-only) — erase & remove-person pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu pip install --no-deps simple-lama-inpainting fire segment-anything
For Linux system packages, the optional AI models, desktop launcher, and platform notes see docs/installation.md.
image-editor is open source and welcomes contributions of any size.
Something crash or render wrong? Open an issue with steps to reproduce and your OS / Python version.
Open an issue →Want a new tool or look? Open a feature request and describe what you'd use it for.
Request a feature →Fork, branch, and send a PR. Read docs/architecture.md first — it maps the modules and the render pipeline.
Fork on GitHub →Docs are plain Markdown in the docs/ folder — easy to edit and submit.
image-editor is built and maintained as a side project. It's free and open source — but development tools, time, and motivation all have a real cost.
If it saves you from a subscription or from losing your photos' resolution, consider chipping in. Every contribution helps keep it maintained and improving.
All contributions go directly to the developer. No company, no VC behind this.
image-editor is free to use, fork, and modify. No telemetry, no accounts, no cloud — everything runs on your machine. (Optional AI model weights carry their own licences.)
View on GitHub →Escanea el QR con tu app bancaria para enviar tu aporte.
¡Muchas gracias!