Full Auto Scanner Conversion
Moderator: peterZ
Full Auto Scanner Conversion
Idea for full auto scanning using a Standard or Hackerspace scanner framework. Book pages on one side are bent back and fanned. Single pages are released by mechanically pulling a hold-down across the fanned pages. An air blast pushes the released page flat against the platen for photographing. Another air blast forces the page over to the other side of the book. The following pics show stages in the page flipping and scanning cycle. This is my original idea and anybody can use it for free.
-
- Posts: 2
- Joined: 30 Nov 2012, 09:31
- E-book readers owned: Kindle 3
- Number of books owned: 0
- Country: Bulgaria
Re: Full Auto Scanner Conversion
Your ideas are just amazing! Can you explain a bit more the mechanical pulling? How will you make sure it releases only a single page? How are you planning on moving the platen?
Re: Full Auto Scanner Conversion
Sorry the drawings are very incomplete. A small gearmotor turning a lead screw could slowly retract the hold-down plate. I don't know much about proximity sensors but I assume some type would work to detect when a single page is released and thereby stop the gearmotor. Maybe somebody with sensor expertise can advise?Atlantius wrote:Your ideas are just amazing! Can you explain a bit more the mechanical pulling? How will you make sure it releases only a single page? How are you planning on moving the platen?
I'm thinking about moving the platen with winch and cable run by a gearmotor.
Thanks for the interest.
- daniel_reetz
- Posts: 2812
- Joined: 03 Jun 2009, 13:56
- E-book readers owned: Used to have a PRS-500
- Number of books owned: 600
- Country: United States
- Contact:
Re: Full Auto Scanner Conversion
Damn, lookin' good! There are plenty of ways to detect a page being turned or not - retroreflective sensors are one, a laser beam-break is another... the hard thing is not so much sensing the page turn. It's turning the page at all. Or, it's doing something once you know you haven't turned the page (or you've turned too many pages).
This is rad, jck57. Looking forward to more. I get a lot of energy just looking at your posts, which is good, because I don't have a lot of energy to spare lately.
This is rad, jck57. Looking forward to more. I get a lot of energy just looking at your posts, which is good, because I don't have a lot of energy to spare lately.
Re: Full Auto Scanner Conversion
Thanks, Dan. Don't work too hard.daniel_reetz wrote:Damn, lookin' good! There are plenty of ways to detect a page being turned or not - retroreflective sensors are one, a laser beam-break is another... the hard thing is not so much sensing the page turn. It's turning the page at all. Or, it's doing something once you know you haven't turned the page (or you've turned too many pages).
This is rad, jck57. Looking forward to more. I get a lot of energy just looking at your posts, which is good, because I don't have a lot of energy to spare lately.
I'm gonna try to build this thing. Just ordered a small gearmotor. Below is a sketch of what I'm thinking. the "shoe" is what I'm calling the block that the pages are bent over. As of right now, I think the shoe will be the movable part that pushes against the stack to slowly release one page at a time. I think this could be a very dependable way to pick up a single page. What I need is a sensor that will tell the Arduino that a page has come off the stack so the Arduino can tell the gearmotor to stop pushing the shoe. IR sensors seem to be popular with Arduino projects but I don't know if IR is good for this.
Re: Full Auto Scanner Conversion
How about a laser trigger setup with Arduino?
Re: Full Auto Scanner Conversion
It's an interesting approach and I look forward to seeing your progress as you prototype it. There are two issues that I can think of that you'll have to be aware of when using the page turning system that you've described:
1) Some books have uneven page widths. It won't take much of a difference in page width to cause multiple pages to be flipped out into the open. If you put enough pressure on the shoe to push out a slightly wider page, the page(s) underneath that might not be as wide could flip out as well.
2) Some books have pages with varying thicknesses. Books with picture sections on slick paper compared to the pages containing only text. As mentioned in the previous point, this would require different pressure put on the "shoe" depending on the change in page thickness.
The second case is more rare than the first, but in both cases you'd want some way to know that these mistakes happened before you unload the scanned book from the scanner.
I would guess that the pressure you put on the shoe to cause one page to flip would vary depending on the number of pages that are bent in that unscanned side of the book. Would you need to worry about different shoe profiles for the largest vs. smallest page width you're planning to scan?
I'm still a fan of pulling a page up with a vacuum to turn it. I think the trick with getting this to work is holding the page down in the middle and slowly rotating the vacuum head as it lifts the page near the outer margin so that it imparts a curl in the page, thereby discouraging more than one page from lifting. It's the same thing your thumb does when lifting/turning a page with your index finger. The thumb holds down the page and the index finger lifts the corner, pulling it back and imparting a curl upward This keeps the pages underneath from coming up with the turned page.
1) Some books have uneven page widths. It won't take much of a difference in page width to cause multiple pages to be flipped out into the open. If you put enough pressure on the shoe to push out a slightly wider page, the page(s) underneath that might not be as wide could flip out as well.
2) Some books have pages with varying thicknesses. Books with picture sections on slick paper compared to the pages containing only text. As mentioned in the previous point, this would require different pressure put on the "shoe" depending on the change in page thickness.
The second case is more rare than the first, but in both cases you'd want some way to know that these mistakes happened before you unload the scanned book from the scanner.
I would guess that the pressure you put on the shoe to cause one page to flip would vary depending on the number of pages that are bent in that unscanned side of the book. Would you need to worry about different shoe profiles for the largest vs. smallest page width you're planning to scan?
I'm still a fan of pulling a page up with a vacuum to turn it. I think the trick with getting this to work is holding the page down in the middle and slowly rotating the vacuum head as it lifts the page near the outer margin so that it imparts a curl in the page, thereby discouraging more than one page from lifting. It's the same thing your thumb does when lifting/turning a page with your index finger. The thumb holds down the page and the index finger lifts the corner, pulling it back and imparting a curl upward This keeps the pages underneath from coming up with the turned page.
Re: Full Auto Scanner Conversion
An idea for a page turning detector with arduino can be seen here. They also made a mockup.
Re: Full Auto Scanner Conversion
Thanks for the comments.
Martin: I don't know if a laser sensor will work because i think you need both an emitter and a target and I think I'll only have one thing to sense proximity. Thanks for the link to the short videos. Unfortunately they don't give much info on what they are using.
Dpc: Not sure what you mean about book widths but I do have an plan to adjust for book width variation. Sorry for the very simplistic drawings so far. My little gearmotor was delivered today. I'm working in the shop today and hopefully will soon have more to share.
Martin: I don't know if a laser sensor will work because i think you need both an emitter and a target and I think I'll only have one thing to sense proximity. Thanks for the link to the short videos. Unfortunately they don't give much info on what they are using.
Dpc: Not sure what you mean about book widths but I do have an plan to adjust for book width variation. Sorry for the very simplistic drawings so far. My little gearmotor was delivered today. I'm working in the shop today and hopefully will soon have more to share.
Re: Full Auto Scanner Conversion
In experiments, the force of sliding the bent pages past the fixed plate caused friction which in turn made the pages release in unwanted ways. This idea is an attempt at improvement. Instead of sliding the pages past the plate, a flexible belt stays in constant contact with the bent pages and is gradually pulled back, releasing a single page at a time. There is thus no relative movement between the belt and the book except for the last few pages.