3 ms·

I worked on this for the library of Uni Würzburg when I was a student. Pretty cool to see it on HN. We worked on high quality scans, processing Terabytes of ra

by pvorb 3y ago

I worked on this for the library of Uni Würzburg when I was a student. Pretty cool to see it on HN. We worked on high quality scans, processing Terabytes of raw image material using Akka (this was around 2012 I think) and also created pipelines for performance OCR at scale on these scans. Doing this on Fraktur and medieval minuscule scripts was tricky and we didn't get really good results during my time there.

pvorb 3y ago
Another thing I worked on was creating interesting visualizations of those books, e.g. http://vb.uni-wuerzburg.de/ub/lskd/regal.html http://vb.uni-wuerzburg.de/ub/lskd/regal.html It was first created using Adobe Flash and then later ported to JS with WebGL/Three.js. God, this looks silly in retrospect.