To bypass GhostScript I wrote a script to extract JBIG2 images directly out a PDF then convert these into PNG using jbig2dec application (this is much faster than using gs).
However, these files have some white lines going through them, whereas they do not have this when they are extracted into PNGs using GhostScript.
I assume this must be a problem with the jbig2dec decoder? Has anyone heard of this issue before?
What decoder does GhostScript use for extracting the JBIG2 images then?
