Digitizing a book with a camera

Magyar leírás:
http://wilika.freeblog.hu/archives/2008/03/02/Digi/

This is how I've "scanned" a book (Mechanics for engineers) to share it with my classmates. (I'm speaking Hungarian, so here's how it went)

I needed: A tripod (a cheap Hama in my case), a camera (Fuji Finepix 6500fd), an old fishtank (my turtles, then, my african snails used to live in it), some fasteners (or what), ACD see, Photoshop, Acrobat

I've pressed the pages against the inner surfaces of an old fishtank (notice how I've fixed it, preventing it from slipping down), so they went flat.
The camera's focus was set on manual, so I've only had to set it once correctly, making the procedure fast.
After taking all sides in I've made an action in photoshop, to enhance the image quality (levels, britghness, contrast, etc. making the old yellowish paper white, and the letters nice and sharp). Since I only had one lamp, I had to apply the adjustments, with a gradient mask, 'cause the top side of the pictures were closer to the light, they were brighter. After the action was finished, I made Photoshop batch all the pictures (est. 200) with it. After it, I've selected every second page, to rotate them.
Then I've combined them in Acrobat, with a few clicks. :)

Category:

Howto & Style

Tags:

• scanning
• books
• book
• how
• to
• digitizing

[ontopofthe7thhill]

Is this a joke?

How much memory is that going to take, let alone the many intricate steps to take the photos store them and well, try to convert them into text. Besides I don't think that's possible. :p

[wilika]

@ontopofthe7thhill

Why would it be a joke? Memory? The time this video was made, my cam had only a one gig card, and my laptop had also one gig ram. Shooting the pages didn't take so long, and the batch process (automated, in photoshop) to enhance the pictures went quick aswell. Since this book had a lot of mathematical texts, and graphs we didn't bother to make a recognition on it. The goal was to make the book available for my mates on the university.

Anyway, Adobe Acrobat X is great in OCR.

[dying2l]

have you figured out how to convert the photo to text?

[wilika]

@dying2l

Adobe Acrobat X offers a great OCR solution: I takes the pages, recognises the texts, and converts them, leaving every unrecognisable part as image. What you get: A pdf, much more smaller then the full-image original, and no information-loss. (and ofcourse a searchable text)

[WalterZelhofer]

Jo otlet, det a Snapscan az gyorsabb (bosc, nem tudok jol irni Magyarul)

[wilika]

@WalterZelhofer

Rákerestem, de annyira nem értem, mi az. Többnyire lapbehúzós (sheet fed?) scannereket találtam erre a szóra.

[wilika]

@lucirz

As far as I remember, it was set to 3 mpx. (the model itself can shoot pictures up to 6mpx)

[lucirz]

@wilika how many Mega pixel is your camera?

[curiousjapan]

it is a shame you do not have two cameras... you could wrap the book around the corner of the fishtank (on the outside of the glass), and shoot the left and right page at the same time

[wilika]

how do you mean it?