Draw custom images on your Revolut card using your phone's drawing screen. RevoDraw automatically detects the drawing boundaries and lets you upload, edit, and draw images via ADB.
- Automatic boundary detection - Detects the dotted-line drawing area from your phone screen
- Multiple image layers - Add multiple images and position them independently
- Edge detection - Multiple algorithms (auto, edges, contours, adaptive) to extract drawable paths
- Real-time preview - See exactly what will be drawn before executing
- Transform controls - Scale, rotate, flip, and position your images
- Eraser tool - Remove unwanted paths with precision
- Reprocess - Adjust edge detection settings and reprocess without re-uploading
- Python 3.8+
- Android phone with USB debugging enabled
- ADB (Android Debug Bridge) installed
- USB cable
Optional: scrcpy if you want to mirror your phone screen on your computer while drawing (not required - you can just watch your phone directly).
You must enable "USB debugging (Security settings)" in Developer Options, then reboot. This allows ADB to simulate touch input.
# Clone the repository
git clone https://github.com/K53N0/revodraw.git
cd revodraw
# Create virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies
pip install -r requirements.txtpython revodraw.pyThen open http://localhost:5000 in your browser.
Steps:
- Connect your Android phone via USB
- Open Revolut app → Cards → Design card → Freeform drawing
- Select the pen tool and set it to the smallest size for optimal results
- Click "Detect Area" in RevoDraw to capture the drawing boundaries
- Upload an image and adjust edge detection settings
- Position and transform your image using the preview
- Click "Start Drawing" to draw on your phone
Tip: Using the smallest pen size gives the cleanest, most detailed results.
Warning: Keep some margin from the detected boundaries - they're not always 100% accurate. If your image is placed too close to the edges, the drawing strokes may accidentally interact with UI elements in the Revolut app and disrupt the drawing process. If this happens, restart and use a slightly smaller scale or reposition the image away from the edges.
Draw shapes:
python revolut_draw.py --heart
python revolut_draw.py --star
python revolut_draw.py --text "HELLO"
python revolut_draw.py --demoDraw images:
python image_draw.py logo.png
python image_draw.py photo.jpg --method edges
python image_draw.py drawing.png --preview # Preview without drawing- Screenshot capture - Takes a screenshot via ADB
- Boundary detection - Uses OpenCV to find the dotted-line drawing area
- Edge extraction - Converts your image to drawable paths using edge detection
- Path scaling - Fits paths within the detected boundaries
- ADB drawing - Sends swipe commands to draw each path segment
| Method | Best For |
|---|---|
auto |
Automatic selection based on image |
edges |
Photos, complex images |
contours |
Logos, drawings (dark on light) |
contours_inv |
Light on dark images |
adaptive |
Images with varying lighting |
- Threshold (0-255): Controls contour detection sensitivity
- Simplify (0-10): Higher values = fewer points, faster drawing
For best results, use images with clean lines and high contrast. You can use AI image generators to create custom artwork that's optimized for RevoDraw.
Use this prompt with your preferred AI image generator:
Create a minimalist line art illustration filling the entire canvas.
Requirements:
- Black lines on white background (or white on black)
- Clean, bold outlines with no gradients or shading
- No fill colors, only strokes/outlines
- High contrast, vector-style artwork
- Consistent line thickness
- Simple but striking design - avoid small intricate details
Theme: [YOUR THEME - e.g., "cyberpunk skull", "geometric wolf", "coding/hacker aesthetic"]
Style: Technical blueprint / line drawing / single-color stencil art
Important: The design must work as a simple outline that could be traced.
Avoid halftones, dots, textures, or soft edges.
For best results, provide the AI with a screenshot of your detected drawing area:
- Click "Detect Area" in RevoDraw
- Take a screenshot of the preview showing the L-shaped drawing boundary
- Send the screenshot to a multimodal AI with this prompt:
I want to draw a custom image on my Revolut card using an automated tool.
The attached screenshot shows the available drawing area (the L-shaped region
with two exclusion zones in the corners).
Generate an image that:
- Fits within this L-shaped boundary
- Uses only black lines on white background
- Has clean, bold outlines (no gradients, shading, or fills)
- Is simple enough to be drawn with continuous strokes
- Theme: [YOUR THEME]
The image will be converted to paths using edge detection, so clean lines are essential.
- Request "line art", "stencil", or "vector style"
- Avoid: gradients, shading, photorealistic styles, fine details
- Best styles: tribal art, geometric patterns, minimalist logos, blueprint drawings
- If the result has too many details, ask the AI to "simplify" or make it "bolder"
"No ADB device connected"
- Ensure USB debugging is enabled
- Check
adb devicesshows your phone - Try different USB cable/port
Drawing doesn't appear on phone
- For Xiaomi: Enable "USB debugging (Security settings)" and reboot
- Make sure you're on the freeform drawing screen
- Try running
adb shell input tap 500 500to test
Boundary detection fails
- Ensure the card drawing screen is fully visible
- Avoid screen overlays or notifications
- Try adjusting screen brightness
revodraw/
├── revodraw.py # Main web UI
├── detect_drawing_area.py # Boundary detection module
├── revolut_draw.py # CLI for shapes/text
├── image_draw.py # CLI for images
└── requirements.txt # Python dependencies
MIT License - feel free to use and modify.
Contributions welcome! Please open an issue or PR.
This tool is for personal use. Use responsibly and in accordance with Revolut's terms of service.