A Python module and command-line tool that generates a Get Stick Bugged Lol video from any image.

gsbl python.jpg python.mp4 -r 760 475 -s 0.7

Example video with sound

This script uses pylsd-nova to detect line segments in the image, Pillow to draw the lines as they move to form the stick bug, and MoviePy to create the video.

• Python 3.6 or later (any OS)

This package can be installed using pip:

pip install get-stick-bugged-lol

Installing the package will register the gsbl command in the terminal (or you can use python -m gsbl). To use the image input.png to generate the video output.mp4:

gsbl input.png output.mp4

Optional arguments:

• -h, --help show this help message and exit
• -r --resolution WIDTH HEIGHT width and height of the video (default: 720 720)
• --img-bg-color R G B RGB background color while the image is visible (default: 0 0 0)
• --line-color R G B RGB color of line segments (default: 255 255 211)
• --line-bg-color R G B RGB background color after image disappears (default: 125 115 119)
• -s --scale SCALE the image scale passed to the line segment detector. Slightly lowering this may improve results in large images. This does not affect the image scale in the video (try --resolution instead). (default: 0.8)

from PIL import Image
from gsbl.stick_bug import StickBug

# Create the StickBug object
sb = StickBug(Image.open('example.png'))  # parameter can also just be a filepath

# Change some properties if you want
sb.video_resolution = (1280, 720)
sb.lsd_scale = 0.5

# That's it! The video will be generated the first time you access it
video = sb.video  # MoviePy VideoClip

# Or you can just save it
sb.save_video('example.mp4')

# If any settings were changed, the video will be regenerated the next time you access it.
sb.line_color = (128, 0, 255)
video_purple = sb.video

• image the source PIL Image. You can set this when initializing StickBug, or at any time by accessing the property. If you want, you can leave this parameter empty while initializing.

• segments a numpy array of the 9 line segments detected. If the line segment detector hasn't run yet, that's done the first time this is accessed. The line segment detector will run again if any other properties have changed. This can also be set manually if you want. Each row of the array must contain the values [x1, y1, x2, y2, width].

• video (readonly) the MoviePy VideoClip generated by the script. If the video hasn't been generated yet, that's done the first time this is accessed. The video will be regenerated if any other properties have changed.

• video_resolution the resolution of the video as a tuple

• lsd_scale the image scale passed to the line segment detector. Slightly lowering this may improve results in large images. This does not affect the image scale in the video.

• img_bg_color the background color of the video while the image is visible

• line_color the color of the line segments in the video

• line_bg_color the background color of the video after the image disappears

Install the required packages:

pip install -r requirements.txt

Install this package directly:

cd [path to repository]
pip install .

Or build a wheel:

cd [path to repository]
python3 setup.py bdist_wheel

This package is available under the MIT License. See LICENSE for more info.

This package makes use of the following external libraries:

• pylsd-nova
• NumPy
• Pillow
• MoviePy