This is a wrapper around the Stanford ljpeg code, which is public domain and also included. It's loosely based on @aaalgo/ljpeg, updated for Python 3 and slightly refactored.
You probably would want to use it to open images from the DDSM Mammography Database, as they had the genius idea of using this near-extinct image format.
You should have NumPy installed:
pip3 install numpyClone this repository:
git clone https://github.com/olieidel/ljpeg3.gitBuild the ljpeg stuff. Some warnings may appear, ignore them and hope for the best. I have no idea how to fix the C code if it doesn't work for you.
cd ljpeg3/jpegdir
makeIn your python project, add ljpeg3 to your path and import it:
import sys
sys.path.append('/path_to/ljpeg3')
from ljpeg3 import ljpegRead a .ljpeg file. There should also be an .ics file in the same directory.
im = ljpeg.read('path_to/some_file.ljpeg')Use NumPy for further processing. Good luck!
In most cases, the height and width information from the ljpeg file is swapped and the ics file is needed to correct it. Oh, and swapped height und width is not equivalent to simply transposing the matrix.. :)
If you don't have an .ics file and your image looks weird, try this:
im = ljpeg.read('path_to/some_file.ljpeg')
im = im.reshape((im.shape[1], im.shape[0])) # swaps width and height
# im = im.T <-- this won't workMIT