Page 1 of 1

marker-crop: Automated, Reliable Postprocessing

Posted: 29 Jan 2017, 22:33
by duerig
One of the biggest problems with book scanning as it is now is the time-consuming and cumbersome postprocessing. I've been working on a way to provide a basic postprocessing workflow that is completely automated with no human intervention.

In order to do this, a user places markers on top of the platen as they scan to indicate where to crop the book and what the orientation of the book page should be. Since the markers are physical tokens, the user can see visually what will be cropped and adjust the markers as they scan.

Right now, I'm beta testing this software and if it proves to be as reliable and automatic as I want, I'll roll it into Pi Scan to make it easier to use. But you can try it out right now if you have a Linux or Mac machine and can install OpenCV. See the project page at:

https://github.com/Tenrec-Builders/marker-crop

It will always be the case that for some people and some projects more complicated post-processing is necessary. But my goal is to make at least the most basic kind of post-processing completely painless with no need to review the book hunting for problems.

Please give me any feedback you have. Especially if you find situations where it doesn't work.

-Jonathon Duerig

Re: marker-crop: Automated, Reliable Postprocessing

Posted: 01 Jan 2018, 10:53
by qqmxdpo
Hi,Duerig.I have written a Script to do same thing .
:D
viewtopic.php?f=20&t=3473

Re: marker-crop: Automated, Reliable Postprocessing

Posted: 16 Jan 2018, 21:09
by gutto12
Does it work for just one camera scans?

tks
Gustavo