mahotas-features.py

New in version 1.4.0: The mahotas-features.py script appeared in version 1.4.0 (July 2015)

With the installation of mahotas, a script called mahotas-features.py is installed, which can be used to compute features from a set of files.

Usage

You call the script with a set of flags specifying which features you want to compute, followed by a list of files. For example:

$ mahotas-features.py --haralick --lbp image-file1.tiff image-file2.tiff

This will output to the file features.tsv (this default can be changed with the --output option)

Full Usage Information

You can obtain help on all the options by running mahotas-features.py --help:

If you use mahotas in a scientific publication, please cite
    Coelho, LP (2013). http://dx.doi.org/10.5334/jors.ac


usage: mahotas-features.py [-h] [--output OUTPUT] [--clobber]
                           [--convert-to-bw CONVERT_TO_BW] [--no-color]
                           [--haralick] [--lbp] [--lbp-radius LBP_RADIUS]
                           [--lbp-points LBP_POINTS]
                           input_file_name [input_file_name ...]

Compute features using mahotas

positional arguments:
  input_file_name       Image files names

optional arguments:
  -h, --help            show this help message and exit
  --output OUTPUT       Output file for feature files
  --clobber             Overwrite output file (if it exists)
  --convert-to-bw CONVERT_TO_BW
                        Convert color images to greyscale. Acceptable values:
                        no: raises an error (default) max: use max projection
                        yes: use rgb2gray
  --no-color            Do not print in color (for error and warning messages)
  --haralick            Compute Haralick features
  --lbp                 Compute LBP (linear binary patterns) features
  --lbp-radius LBP_RADIUS
                        Radius to use for LBP features
  --lbp-points LBP_POINTS
                        Nr of points to use for LBP features