[dahart]

> While there are some computer algorithms that can help dewarp the pages after capture, it is always more reliable to just capture flat pages in the first place.

I’m sure this is technically true, but curious how much it matters in practice today? Reading Google’s book scanning patents I found a description of a de-warper based on capturing a 3d depth scan of the book, which I assumed they were using in order to achieve the scale of scanning all books on earth. Capturing and de-warping a 3d depth scan would also be leagues more reliable than trying to do a purely 2d image based de-warp.

> The lights must also be positioned to minimize glare and reflections.

For my personal photo scanning and archiving project, I used a polarizing filter on the light and on the camera in order to eliminate specular glare, it works amazingly well. Would that be impractical, and/or not work as well on books for some reason?