Re: [Denemo-devel] SharpEye, PhotoScore, and Denemo

[Nathan Stewart]

I've always been puzzled that musical OCR is so difficult. It seems like a VERY constrained problem set compared to text. Small alphabet. Somewhat regular rules Not even a lot of typefaces.. Sure handwritten would be hard, and some advanced things might be difficult to detect/codify, but basic SATB, nothing more complex than 8th notes, ties, and repeats seems like it should be very straight forward.

On Thu, Dec 15, 2016 at 10:21 AM, Johan Vromans <address@hidden> wrote:

On Thu, 15 Dec 2016 14:29:45 +0000, Richard Shann <address@hidden>
wrote:

> If you do try these OCR programs out I would be *very* interested in the
> MusicXML generated for the attached PDF ... I tried Audiveris recently
> on this (that program - Audiveris - is supposed to be imminently getting
> an upgrade ... has been for months ...) to no very good effect.

Results using SharpEye attached.

AllPages.xml : MusicXML produced by SharpEye
AllPages.ly:   result from musicxml2ly  --npl --nsd --nb
AllPages.pdf:  lilypond AllPages.ly

Time spent: 4 minutes.

-- Johan
_______________________________________________
Denemo-devel mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/denemo-devel