06-07-2021, 02:23 PM
(This post was last modified: 07-07-2021, 01:42 PM by Mark Knowles.)
In an ideal world I would like to be able to highlight every molecule of ink and wipe everything else such as imperfections in the vellum, creases, stains etc. Now obviously that not going to happen, but it does say to me that despite the amazingly good quality scans of the Voynich that we have access to, they are far from perfect.
At the moment it is often unclear what is ink and what is not. Distinguishing between the two would help unravel the pen path that produced the writing on the vellum.
I wonder what scope there is for the use of software to transcribe the Voynich. Maybe a machine would be better at distinguishing the characters. Obviously handwriting recognition software exists, though is limited, for the latin alphabet. A machine could do things that people can't like cross compare all instances of a character as an aid to identification. Possibly a human could assist the algorithm by doing things like identifying where the text is, though I think Voynichese.com already does this.
At the moment it is often unclear what is ink and what is not. Distinguishing between the two would help unravel the pen path that produced the writing on the vellum.
I wonder what scope there is for the use of software to transcribe the Voynich. Maybe a machine would be better at distinguishing the characters. Obviously handwriting recognition software exists, though is limited, for the latin alphabet. A machine could do things that people can't like cross compare all instances of a character as an aid to identification. Possibly a human could assist the algorithm by doing things like identifying where the text is, though I think Voynichese.com already does this.