austinschwartz / hatched

Convert images to plotter-friendly hatched patterns

hatched

Library and vpype plug-in to convert images to plotter-friendly, hatched patterns.

Built with OpenCV, scikit-image, Shapely, matplotlib and svgwrite. You can reach the author Drawingbots's Discord server.

Getting Started

Using with vpype

Using hatched as a vpype plug-in is the easiest way to get started. For that, simply install hatched with the following command, ideally in a virtual environment:

$ pip install git+https://github.com/abey79/hatched.git#egg=hatched

vpype is automatically installed with hatched, so no further steps are required.

You can confirm that the installation was successful with the following command, which also happens to tell you all you need to know to use hatched:

$ vpype hatched --help
Usage: vpype hatched [OPTIONS] FILENAME

  Generate hatched pattern from an image.

  The hatches generated are in the coordinate of the input image. For
  example, a 100x100px image with generate hatches whose bounding box
  coordinates are (0, 0, 100, 100). The `--scale` option, by resampling the
  input image, indirectly affects the generated bounding box. The `--pitch`
  parameter sets the densest hatching frequency,

Options:
  --levels INTEGER...             Pixel value of the 3 thresholds between
                                  black, dark, light and white zones (0-255).
  -s, --scale FLOAT               Scale factor to apply to the image size.
  -i, --interpolation [linear|nearest]
                                  Interpolation used for scaling.
  -b, --blur INTEGER              Blur radius to apply to the image before
                                  applying thresholds.
  -p, --pitch LENGTH              Hatching pitch for the densest zones. This
                                  option understands supported units.
  -x, --invert                    Invert the image (and levels) before
                                  applying thresholds.
  -c, --circular                  Use circular instead of diagonal hatches.
  -d, --show-plot                 Display the contours and resulting pattern
                                  using matplotlib.
  -l, --layer LAYER               Target layer or 'new'.
  --help                          Show this message and exit.

To create a SVG, combine the hatched command with the write command (check vpype's documentation for more information). Here is an example:

$ vpype hatched --levels 64 128 192 -s 0.5 -p 4 input.jpg write --page-format a4 --center output.svg

Using hatched as a library

To play with hatched, you need to checkout the source and install the dependencies in a virtual environment, for example with the following steps:

$ git clone https://github.com/abey79/hatched.git
$ cd hatched
$ python3 -m venv venv
$ source venv/bin/activate
$ pip install -r requirements.txt

Running the example

Example can then be run by executing the corresponding file:

$ cd examples
$ python skull.py

The processing result is displayed in a matplotlib window:

A skull.svg file is also created with the output graphics.

Usage

Call the function hatched.hatch() to process your image. It takes the following parameters:

• file_path: input image (most common format are accepted)
• image_scale: scale factor to apply to the image before processing
• interpolation: interpolation to apply for scaling (typically either cv2.INTER_LINEAR or cv2.INTER_NEAREST)
• blur_radius: blurring radius to apply on the input image (0 to disable)
• hatch_pitch: hatching pitch in pixel (corresponds to the densest possible hatching)
• levels: tuple of the 3 thresholds between black, dark, light and white (0-255)
• h_mirror: apply horizontal mirror on the image if True
• invert: invert pixel value of the input image before processing (in this case, the level thresholds are inverted as well)
• circular: use circular hatching instead of diagonal
• show_plot: (default True) display contours and final results with matplotlib
• save_svg: (default True) controls whether or not an output SVG file is created

License

This project is licensed under the MIT License - see the LICENSE file for details.

The example image skull.jpg is licenced under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.