Hi Lode,

I am not sure I understood your question, I guess you are speaking about scanning books in 2D or in 3D.

If you work in 3D, than you can get orthophotos directly from 3d models and you will have no problems about distortions (to see what I mean you can check this post where you can compare a simple 2D rectified picture VS a real orthophoto obtained from a 3D pointcloud: http://arc-team-open-research.blogspot.it/2013/12/from-drone-aerial-pictures-to-dem-and.html It refers to archaeological remains, but it would be the same for books. Look the slider at the end of the post).

If you work in 2D, you could use simple metric references to rectify the images (you could do this also inside a simple open source GIS like QSGI, in the same way archaeologist do phtomapping). Here an example (http://arc-team-open-research.blogspot.it/2014/02/quantumgis-and-openjump-photomapping.html), but now you can do everything easily in QGIS, without the need to use other software.

If you do not need metric values, but you just want to correct distortions and aberrations in a fast and easy way, than you can just sdjust your pictures within a raster graphic open source software like GIMP, using the "perspective" tool.

I hope what I wrote will be useful. If you need more infos, just ask. Have a nice day!

no I meant aberration in 3D scanning by the structured light system. Identical to the aberration or distortion one gets on the Zeutschel book scanners. I'm using an HP Sprout, but it is nerve wrecking. I'm doing this as part of my thesis and since I have experience with book scanning and with many other kinds of scanning (since the Kurzweil deepfrize size scanners and the HP Scanjet for the Apple Mac fx in the early 1990s).

The structured light 3D scanning creates trapeziums and merging/fusing proves to be an headache. As with lens aberration in Photoshop it shouldn't be that difficult to correct based on software. But the question is HOW?

Hi Lode,

to rectify (and correct) 3D data (if you use 3D structured light scanners), you can use two FLOSS: MeshLab or Cloud Compare. They work with 3D data and can rectify them with known points (e.g. you can use som CGP). Than, if you extract an orthophoto form the 3D models there should not be aberrations. Here an old videotutorial: http://arc-team-open-research.blogspot.it/2014/09/georeference-mesh-using-reference-scene.html This would be the ore accurate and precise way, but for book's pages I guess that, like you say "is an headache". I think for book's pages would be simpler to work in 2D photogrammetry with images from normal scanners like the one you are using.

If you work in 2d, the best way is to use a photogrammetry extension of a simple GIS (a FLOSS one is QGIS). This would be a simple and good enough solution. If you want something simpler (also when less precise and accurate) you could also simply use a normal raster graphic software (e.g the FLOSS GIMP which is like Photoshop) and remove distortions and aberrations. For instance within GIMP you could remove lens distortions with a specific plugin and then, considering the page size, use the "perspective tooll" to adjust aberrations.

If you want, send me the image of a page and its real size. I can try to record a short videotutorial.

Greetings

thanks