https://github.com/piebro/svgdatautil
https://github.com/piebro/svgdatautil
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/piebro/svgdatautil
- Owner: piebro
- Created: 2021-01-04T15:18:43.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2021-01-10T12:02:26.000Z (over 5 years ago)
- Last Synced: 2025-01-18T04:43:32.714Z (over 1 year ago)
- Language: Python
- Size: 264 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Library for usage of SVGs in machine learning
Library to simplify SVGs for machine learning. The simpler SVGs are constist of polylines with discrete points.
## Installation
```bash
pip install -q git+https://github.com/piebro/svgdatautil
```
## Usage
### Simpify SVG
```python
from svgdatautil import svg_to_lines, simplify_lines, save_svg, get_svg_from_lines
# the points of the lines should be in [0 0 1000 1000]
lines = svg_to_lines("examples/1BKZ.svg")
sim_lines = simplify_lines(lines, max_xy=1000, rdp_epsilon=1) # rdp_epsilon describes how much the lines are simplified
save_svg("examples/1BKZ_1000_1.svg", get_svg_from_lines(sim_lines, background=None))
```
### Create simplified SVG dataset
```python
from svgdatautil import svg_dir_to_dataset_pkl, save_simplified_dataset_pkl, dataset_pkl_to_lines
svg_dir_to_dataset_pkl("examples", "examples.pkl")
save_simplified_dataset_pkl("examples.pkl", "examples.pkl_2_500_1.pkl", svg_count_in_new_dataset=2, max_xy=500, rdp_epsilon=1)
svg_lines = dataset_pkl_to_lines("examples.pkl_2_500_1.pkl")
```
## Example
Original SVG with 224K:
`max_xy=10000` and `rdp_epsilon=0.5` with 40KB:
`max_xy=1000` and `rdp_epsilon=1` with 28KB:
`max_xy=250` and `rdp_epsilon=2` with 20KB:
