ShapeFromShading

Converting Images of seals into 3D objects

Part of the Coding Da Vinci Niedersachsen 2020 Hackathon, we scan images of seals and produce 3D object out of them.

Instruction:

You need Python version >= 3.8

Unix

python -m venv venv
. venv/bin/activate
pip install -e . -r requirements.txt
sealconvert3d [your source file]

Windows

python -m venv venv
.\venv\Scripts\activate.bat
pip install -e . -r requirements.txt
sealconvert3d [your source file]

Usage

For development, after clone, run:

pip install -e . -r requirements.txt

Now,

sealconvert3d [args]

should be available in your command line.

Options are:

usage: sealconvert3d [-h] [-o OUTPUT] [-alg {tsai_shah,tsai_shah_specular}] [-s SLANT] [-t TILT] [-i ITERATIONS] [-l LOGLEVEL] source

Seal Shape From Shading out of Potsdam

positional arguments:
  source

optional arguments:
  -h, --help            show this help message and exit
  -o OUTPUT, --output OUTPUT
                        Filename to write to
  -alg {tsai_shah,tsai_shah_specular}, --algorithm {tsai_shah,tsai_shah_specular}
                        Shape from Shading Algorithm to use
  -s SLANT, --slant SLANT
                        Slant of lighting source
  -t TILT, --tilt TILT  Tilt of lighting source
  -i ITERATIONS, --iterations ITERATIONS
                        Iterations to run the algorithm through
  -l LOGLEVEL, --loglevel LOGLEVEL
                        Set debugging level. Use "debug" for highest details

Image Attributions

The images uni_big_greyscale.png and uni_big_heightmap.png are from Susanne Krömker and Mara Hubert, and licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. Further information can be found under here