Daniel Reetz, the founder of the DIY Book Scanner community, has recently started making videos of prototyping and shop tips. If you are tinkering with a book scanner (or any other project) in your home shop, these tips will come in handy. https://www.youtube.com/channel/UCn0gq8 ... g_8K1nfInQ

My 2nd new book scanner!

Built a scanner? Started to build a scanner? Record your progress here. Doesn't need to be a whole scanner - triggers and other parts are fine. Commercial scanners are fine too.
Gandalf
Posts: 17
Joined: 23 Apr 2017, 15:21
Number of books owned: 20
Country: Canada

My 2nd new book scanner!

Post by Gandalf » 28 Jun 2017, 01:25

I wanted to make a new book scanner without an acrylic platen but i found it was hard to make the page even with just two arms. So i changed my plan and the result looks like this.

This one is similar to Mohib's scanner but, two things are different. I thought Mohib's scanner had one big problem. The main problem is that the book is easily movable. The book or platen was is not fixed. I solved this problem by securing platen . Then I wanted to make pages easy to turn. My scanner makes it easy to lift up the platen. But mine is not portable. Mohib's is more better than my scanner for carrying.

https://ibb.co/dOkrtQ
https://ibb.co/iVrmSk
https://ibb.co/dPHRSk
https://ibb.co/eLrmSk
https://ibb.co/cmNRSk

The scanned page is this.
https://ibb.co/mYZvL5

User avatar
Mohib
Posts: 87
Joined: 05 Apr 2014, 21:15
Number of books owned: 0
Country: Canada

Re: My 2nd new book scanner!

Post by Mohib » 28 Jun 2017, 11:10

That's a nicely built scanner (your adjustable, tensioned hinge on which the platen tilts is clever) and I'm glad my design was a useful inspiration.

I have few questions and observations about your design.

a) About your sample page scan ...

In the photos, it looks like you've got an iPhone 4 sitting on top of your scanner, but the EXIF info from that impressive sample page image indicates it was taken by a Samsung Galaxy Note 4. I checked its specs and noted it's got a 16 MP camera, so it's very good to see what it can deliver.

As I suspected, from comparing image quality from smartphone cameras at the site below, newer smartphone cameras should deliver much higher quality images than my old iPhone 4s' 8MP camera, but it's always good to see an actual scan, so thanks for sharing that.

http://www.gsmarena.com/piccmp.php3?idT ... hone3=4212

- What camera software did you use?
- Did you zoom in at all?


b) About the rubber bands over the left hand pages and their supports ...

- Do you find that the rubber band vertical supports interfere with your left hand/arm during page turning? It would seem you'd have to be careful not to hit it with your left hand/arm.

- Was there a reason you decided to use them to hold the left hand pages back rather than using your left hand to hold the pages back?

- Did you consider using a smooth, flat black plastic sheet, that let the book slide easily instead of the "sticky" rubber, shelf-liner under the book? It would seem to me, given your platen is fixed, a slippery under-sheet would let you slide the book so its spine engages the fixed platen would let you, as in my video, use your left hand to hold and turn the pages efficiently as well as eliminate the rubber bands and their supports. There might be a slight up/down drift while scanning, but assuming you're using Scan Tailor to process the images, it will find and select the content.

BTW, I've found that particular style of shelf liner, that you've used (and I originally used), confuses Scan Tailor's content searching algorithms and it thinks all those reflections from the dimples (or light areas underneath showing through the gaps) are part of the page content and sometimes I had manually adjust the content box on every page. That's why I now use a solid black mat underneath.


c) Similarly about the support structure on the right (where the bungee cord attaches) ...

- Do you find it restricts/interferes your right hand/arm movement when tilting the platen up?


d) About the "one big problem" with my design you mentioned :) ...

You felt my "free floating" book/platen design was a "big problem", however, in practise, it's actually not as big a problem as it seems, or even one at all. It only takes scanning a couple of books to find a style that works for you so you're able to generally control the book from drifting and/or restoring its position when needed. And so given this, yes, there is a little "skill" to learn, because in my design the operator (like the table top) is actually a "component" of the scanner, as compared to the more "mechanical" scanner designs that all but remove the person from the equation (save for page turning).

Nevertheless, I have thought about various ways to either fix platen and/or book position, so one would force the registration of the other, but found anything I tried a) increased construction complexity, b) reduced portability (unless you travelled without that extra hardware), c) dramatically reduced scanning speed (because extra hardware breaks a fluid scanning rhythm, as that shown in my video) and d) all for not much, if any, real practical benefit. I therefore deliberately decided to keep the operator as a component of the scanner as it simplified it and let it work better.

The only thing I've not tried to help with book drift so far, that might work but probably only with hardcovers, is some sort of adjustable/fastenable edge against which the right hand edge (i.e. opposite the spine) that the hardback cover on the table can be kept pushed against (i.e. your left hand is pushing the book to the right against the edge). That would fix the book position and then the platen would register in the spine. However, I'm still thinking of a way to do it simply, somehow attached to the vertical to avoid a lot of c-clamps or other hardware.

Gandalf
Posts: 17
Joined: 23 Apr 2017, 15:21
Number of books owned: 20
Country: Canada

Re: My 2nd new book scanner!

Post by Gandalf » 28 Jun 2017, 15:43

First above all, I am sorry for my comment, "one big problem" on your scanner. It is not "one big problem" as you mentioned.
I apologize for the roughness of my expression on that matter.

At first I wanted to make a book scanner without a platen. But the pages were not even than I expected. So i changed plan and realized my scanner would overcome the weak point(?) of your scanner.

A simpler design makes more sacrifice. When we make a book scanner I think we need essential parts necessarily, light, platen, camera, and
cradle. Your book scanner doesn't have cradle part. This means we have to handle the book not to be floated by hands. So that's why I mentioned it. My scanner sacrifices your simplicity instead.

1. You have really good eyes. When I scanned my book I used my galaxy note 4. Later I replaced it with iPhone on the top to take pictures of my rig. If it made you confused I am sorry. I didn't use any other camera app. I just used it's own camera app. The app you suggested is for iOS. I am searching apps to fix focus. My scanner is specialized under letter size (11"×8.5"). So I don't need zoom. The height is fixed. The scanned page is 6 mega pixel size(1836x3264).

2. The elastic band on the left hand is not interfering. It is just holding it's cover or pages. I used my left hand to hold a book at first but because the book was not float and the elastic band was holding the pages I just turned the pages under the platen.

3. The black mat is for acylic platen. To protect the scratch of platen I bought it, cheaper one, at a dollar store. But i didnt know the mat with dimples can confuse scan tailer's content-searching. Thank you for your good comment.

4.The black bungee band is also not interfering. When I push down the platen, my hand is always on the lever side.

When I am scanning a book, my right hand lowers the platen down and turns the pages. With left hand, I keep the book up to platen. This can make scanning slowly because I don't want to move the book. Actually I cannot do fast not because of hindrance of the elastic bands but because of restrictions of power of the bands. I have to allow returning time of bands.
Last edited by Gandalf on 28 Jun 2017, 18:25, edited 3 times in total.

Gandalf
Posts: 17
Joined: 23 Apr 2017, 15:21
Number of books owned: 20
Country: Canada

Re: My 2nd new book scanner!

Post by Gandalf » 28 Jun 2017, 16:16

This is my first scanner that I made without platen. I could adjust the length to press top and boottom of the book.
https://ibb.co/b71X2k
https://ibb.co/bvE8a5

User avatar
Mohib
Posts: 87
Joined: 05 Apr 2014, 21:15
Number of books owned: 0
Country: Canada

Re: My 2nd new book scanner!

Post by Mohib » 28 Jun 2017, 18:41

Gandalf wrote:
28 Jun 2017, 15:43
First above all, I am sorry for my comment, "one big problem" on your scanner. It is not "one big problem" as you mentioned.
I apologize for the roughness of my expression on that matter.
No need to apologise. Every scanner design is a blend of compromises and the strengths and weaknesses of any design really depend on the individual's perspectives and objectives. So, what maybe one person's "big problem" maybe irrelevant to another person.
Gandalf wrote:
28 Jun 2017, 15:43
When we make a book scanner I think we need essential parts necessarily, light, platen, camera, and cradle. Your book scanner doesn't have cradle part. This means we have to handle the book not to be floated by hands. So that's why I mentioned it.
Ah, interesting perspective. From my perspective, a cradle, in the conventional sense, is not an essential, necessary part of a book scanner.

As I saw it, the primary purpose of a cradle is to keep the book in position under the camera and so, in my mind, any method which achieved that objective, or mitigated it to the point that any change in the book's position is minor and resolvable with little to no effort, was sufficient. Thus, the table top itself, a rubber mat, and a little operator skill in handling the book, was sufficient for me.

Typically, the cradle is intimately connected to the platen, fixing platen/camera distance to ensure a consistent image scale, but that is not essential for my scanning objectives. In any event, image scale is no longer a problem with the image scaling/page numbering script (albeit with a little image processing, and the slight image degradation that might bring, which is fine for me, but perhaps not for others).

For me, the primary purpose of the platen is to flatten the page and so any method that could press the book against the table (i.e. the "cradle") was sufficient. My priority was a) a simple method, b) a method that allowed the platen to be removed and returned quickly for fast page turning, and c) a method that worked as well and as fast for tightly bound paperbacks and books with text close to the gutter without damaging the binding from over-spreading it.

Nevertheless, you've got a good adaptation of the original principle of the TIFLIC scanner and I'm sure it will spur more ideas by others looking to experiment further. Looking forward to learning more about how it worked for you.

User avatar
Mohib
Posts: 87
Joined: 05 Apr 2014, 21:15
Number of books owned: 0
Country: Canada

Re: My 2nd new book scanner!

Post by Mohib » 29 Jun 2017, 13:19

I was thinking about your design because I was confused by your comment:
Gandalf wrote:
28 Jun 2017, 15:43
My scanner is specialized under letter size (11"×8.5"). So I don't need zoom.
until I realised something important, which I hadn't thought of before, about all scanners (including most (all?) V platens) that fix the platen/cradle x & y position relative to the camera.

Since the platen's x & y position is fixed relative to the camera, the spine of all books, no matter the size, will appear in the same place on the image. This means the camera zoom must be fixed, just like you said, because if the camera is zoomed in, to enlarge pages of smaller books, the camera will zoom into the centre of the platen and **not the centre of the book's page** -- because the book is forced way over to one side of the platen because the spine has to engage the edge of the platen -- and the page will be severely cropped, or even not even appear in the image at all (if the book is particularly small).

In other words, with fixed platen/camera designs, depending on the book size, **the book shifts left and right** relative to the centre of the image.

And this means for smaller books -- like paperbacks which often have small type -- the full resolution of the camera can't be used to image the full page, which of course reduces the image quality of small pages if you have to enlarge them to see them better and/or also compromises OCR quality of smaller type on smaller pages. So this maybe an important consideration, when choosing a scanner design, if one is scanning a lot of small books with small type (like paperbacks)

I never realised this because in the TIFLIC scanner, since the platen is not fixed relative to the camera, neither is the book. Without thinking, I simply position the book (regardless of size) in the centre of the camera's field of view, zoom in (and/or use the macro rail to move the camera itself -- shown at it's lowest position in the schematics below -- to avoid excessive zoom) and enlarge the page so it occupies as much of the full image size as is reasonable (and, thereby, use the maximum resolution of the camera no matter the page size). Then, I just drop the platen in place -- again without thinking about it -- because its position relative to the camera is unimportant.

In other words, with TIFLIC, instead of the book shifting left and right relative to the centre of the image, as in fixed platen designs, **the platen shifts left and right** depending on the book size.

I've attached 4 schematics below to illustrate this on a TIFLIC scanner. Two show the platen/book/camera field of view (FOV) with a small book and two with a large book. The relevant book size/platen position/FOV are in red and the other size is in blue to show the platen shift. Note that with a small book I mount the platen handle on a different set of holes in the platen so the handle doesn't shift as much as the platen itself.
.
.
TIFLIC Normal Scanner (V2.1) - Platen shift for large & small book - Large - Side - 50%.png
TIFLIC Normal Scanner (V2.1) - Platen shift for large & small book - Large - Side - 50%.png
.
.
TIFLIC Normal Scanner (V2.1) - Platen shift for large & small book - Small - Side - 50%.png
TIFLIC Normal Scanner (V2.1) - Platen shift for large & small book - Small - Side - 50%.png
.
.
TIFLIC Normal Scanner (V2.1) - Platen shift for large & small book - Large - Top - 50%.png
TIFLIC Normal Scanner (V2.1) - Platen shift for large & small book - Large - Top - 50%.png
.
.
TIFLIC Normal Scanner (V2.1) - Platen shift for large & small book - Small - Top - 50%.png
TIFLIC Normal Scanner (V2.1) - Platen shift for large & small book - Small - Top - 50%.png
.

Gandalf
Posts: 17
Joined: 23 Apr 2017, 15:21
Number of books owned: 20
Country: Canada

Re: My 2nd new book scanner!

Post by Gandalf » 29 Jun 2017, 23:06

The enthusiasm for your book scanner makes you realize more. I hope you have more confidence in your scanner.

When I scanned a book with a Canon sx200 I used zooming for a small book. In this case I experienced two problems. First, I had to move the camera to be centralized. So I need a 4-way(up & down, L&R) slider.

If I fix the camera's position I have to move a platen or a book to be in the middle. But if a platen is fixed I have to move the camera to take pictures. The third option is i am scanning a page always in same sizes for any sized books in the fixed camera and platen. I chose this option now. Additionally, a point & shoot camera often returned to it's original condition after short break time when I was using zooming. Danel too suggested it would be better not to use zooming on this problem.

As you mentioned, we need zooming for small books to get more bigger letters for OCR. At first I wanted to make a slider to cover this problem. You can see it in my first scanner without a platen.

But the maximum height of my cell phone camera in my scanner is just 13.5". This means zooming is not a big issue when I considered letter sizes according to the distances. Even though I can zoom in to enlarge the letters in the same height I just shoot without zooming. The letter size is a sacrifice of the fixed platen.

As a matter of fact my camera is easy to move to the left or right to be in the middle of the book because i used a phone case with velcro tape.

I think your scanner can be adjusted easily to the centre for any books because the platen is not fixed. But because of that, the centre can be relative also. Even if it is a minor difference.... Your passion for the scanner now is enlarging your experiences.

Anyway, you said the cradle is not essential. But I don't think so. We need always cradle to adjust the book to be central to the camera. Your hands and desk are doing this. We only replaced it with our Hands to make a scanner more simpler. I mentined "essential parts" in the view of function. Good scanning always needs 4 elements/components. When I have studied all kind of book scanners on this community, every scanner has weak points or strong points according to 4 components. Even a book scanner with 4 elements should instead give up simplicity. However, many people try to make a well-funtioned and simple book scanner. But It is difficult to catch two hares at once. If you check all kind of book scanners you can find it. Without hight technology we have to remove one of than as sacrificial one for simplicity. You can find easily that many over-the top scanners do not have platen or cradle.

My scanner is good for paper books too because the elastic band is holding the pages. And if you attach the steel plate to the left side of the levers you can have almost 2 mm of gutter accoring to thickness of a plate. My left side acylic plate is for the bensing of a platen because the acylic is not hard so it can be bended by pressing power a little bit . The left plate can solve this problem and also keeps it close to the book.

I am not an English speaker. Please understand me for roughness of my expression.

User avatar
Mohib
Posts: 87
Joined: 05 Apr 2014, 21:15
Number of books owned: 0
Country: Canada

Re: My 2nd new book scanner!

Post by Mohib » 30 Jun 2017, 01:40

Gandalf wrote:
29 Jun 2017, 23:06
I am not an English speaker. Please understand me for roughness of my expression.
Your English is fine, and no misunderstandings at all.
Gandalf wrote:
29 Jun 2017, 23:06
The enthusiasm for your book scanner makes you realize more.... Your passion for the scanner now is enlarging your experiences.
Ha ha! Not so much my enthusiasm for my scanner but my interest to better understand the problems of different book scanner designs. I was just pointing out an interesting problem I had not realised until you made your comment about zoom, and so thought it maybe of interest to others who also may not have realised it, so they can understand that most scanner designs do not use the full resolution of the camera for all page sizes.
Gandalf wrote:
29 Jun 2017, 23:06
You can see it in my first scanner without a platen.
Interesting. I had not noticed that before.
Gandalf wrote:
29 Jun 2017, 23:06
Additionally, a point & shoot camera often returned to it's original condition after short break time when I was using zooming. Danel too suggested it would be better not to use zooming on this problem.
Yes that can be a problem with some point and shoots with power-save. But also usually they do not do continuous zoom, but stepped zoom, so the exact zoom setting can be restored.

With smart-phones it's not an issue. However, smartphones use digital zoom so some resolution is lost depending on the amount of zoom used.

So the best solution, to use the maximum resolution of the camera, for point and shoots or smartphones, is to have z-axis flexibility so the camera can be moved up and down towards the book -- as you did in your first design without a platen, and I also do with the macro focusing rail.

I am looking into mirror-less-DLSRs (with electronic shutters, not mechanical which will fail). These allow proper manual control over zoom and other functions and also have the option of using much better lenses for much higher quality images, which is important for some of my scanning applications.
Gandalf wrote:
29 Jun 2017, 23:06
Anyway, you said the cradle is not essential. But I don't think so. We need always cradle to adjust the book to be central to the camera. Your hands and desk are doing this. We only replaced it with our Hands to make a scanner more simpler. I mentined "essential parts" in the view of function.
Ok, so just a different understanding. I was thinking of "essential parts" from the view of construction.
Gandalf wrote:
29 Jun 2017, 23:06
My left side acylic plate is for the bensing of a platen because the acylic is not hard so it can be bended by pressing power a little bit .
Ok that explains that part. I was wondering why you had it. FYI, I used 3/8" Plexiglas and it doesn't bend.

dpc
Posts: 270
Joined: 01 Apr 2011, 18:05
Number of books owned: 0
Location: Issaquah, WA

Re: My 2nd new book scanner!

Post by dpc » 30 Jun 2017, 14:30

If you're using a lightweight camera like a smartphone, why wouldn't you just mount the camera to the platen and then get rid of that tubing and camera mount hardware?

If you did this, you won't have to worry about moving/shifting books or changing scale factors at all.

Gandalf
Posts: 17
Joined: 23 Apr 2017, 15:21
Number of books owned: 20
Country: Canada

Re: My 2nd new book scanner!

Post by Gandalf » 30 Jun 2017, 16:41

What is "tubing"? You mean "steel round bar" or left side handle?

If I install the camera on the platen, the camera can be shaken even though it is tightened. This means you could get a blured page. As you know, the more faster, the more shaking! I wanted to reduce the vibration as much as possible. Second, if i mount a camera on it, it can be more heavier and slower to lift and lower. Third, I need to install a lamp at top right corner. A camera mount can hold it. Finally, if i install it on the platen, the construction can is more complicated. So I need to change a design for the platen's levers. Perhaps that is unlikely to be simplistic. For these reasons, it is better for me to make it separately.

Post Reply

Who is online

Users browsing this forum: No registered users and 7 guests