Hmm ocr might not be the best way to get the value of a book cause of editions. IE "the dragon book" the first edition is worth alittle bit but the other editions arnt worth as much, along with some books with the same name it might pull false positives. But i guess a 2 stage system might work. Ocr as a preliminary scan that tells you the name of the book to scan the imei number to double check to make sure it's the same book found

@jckimble I was imagining there might be some visual distinctions between editions that could help classify, but you're right, edition and condition play a huge part in valuation

@trogus there are visual distinctions between editions usually like in my example above 1st edition is called the red dragon and 2nd edition is called the purple dragon which is easy to tell the difference between looking at it but even the purple dragon is $170 in used condition so might have been a bad example.

But if you know of a few apis that have imei numbers or book values i would be willing to help on this idea since it would give me something to do at yard/estate sales when my wife drags me to them

Considering the amount book there is out there, it might close to impossible. But great idea!

@tahnik @trogus You should do a competition with book covers for your beta launch.. But that would result in a miserable 5-10% of the total books

You could add an ISBN scanner, that would be enough for most modern books, and easier for the user than a whole cover picture.

Goggles from Google can kinda do this already. It will identify the scanned object, in your case a book and do some googling on it. They even have an example on the play store of identifing a book.

@jckimble I think you mean ISBN.

@jckimble
Amazon (Paid; Unlimited): http://docs.aws.amazon.com/AWSEComm...
Google (Free; Courtesy Limit: 1000/day): https://code.google.com/apis/...
ISBNdb (Free; Rumoured limit: 500/day): http://isbndb.com/api/v2/docs

I love the idea because I am an avid book lover, hauler, reader, and collector...but I am sadly concerned and aware of the constant demise of bookstores, both corporate and local mom and pop shops; ugh it makes me sad.

Cool concept, but I am not sure there would be a demand for people to actually use it.

@unjx I agree, the market size for such a niche problem might only be in the dozens of people. But think of the glory!

You could look into Vuforia

Vuforia (as suggested), wikitude, aurasma, and more. Like...a lot. It depends on where you are gonna distribute it. And how. Oh, and if you are interested in code or just a user.

Vuforia works with unity if you are interested in coding.

@trogus market? Keep the app for yourself and make millions acquiring and selling rare books!

Holy shit, I had an idea like this over a year ago! My idea was an app that scans ISBN barcodes (or manually enter ISBN / other details). Then it would look up on Amazon, eBay, etc to see how much other people are selling it for.

Never got around to making it, but even if I did, I'm not sure I'd share it ;)

If you want help, I'd love to help you out! I recently worked on an AR program for a hackathon. 😁

The American Pickers and Pawn Stars would love such a technology, but of course they'd purchase it for cheap 😂

As an often ar developer and knowing the ins and outs of vuforia, wikitude, tango sdk, and al ton of other high profile image/room/etc. Recognition sdks.
I sadly think your project is impossible with current standards of AR.

First of all most older books lack destinction in patterns and most ar tools are practically colorblind so that will not help either.

Then you want to recognize the book as in a stack or in a row of books, giving you even less data to go on.

And then you need accurate target graphics of all those rare books to even be able to recognize it in the first place.

Conclusion: what you can do is make a database of rare books, use mainly book text to be able to search it. And of every book list what makes it a valid and expensive version and what features it should or should not have.

In the end you could expand beyond books (think rare LP's? Etc.) And make the app a bit more high profile. Consider integrating with existing website that already have this data.

Sorry!

How do you plan on finding and storing individual book identity like an ISBN or OCR. Is there an existing database for that?

Interested in joining the team mail me
rahuljaisinghani1997@gmail.com

Future market :)

@teganburns I can just talk for myself but I was thinking about something like that on my last visit to a flea market. Sounds like an interesting project!

Can I join?

It’s a deal for ARKit. But the problem is a book covers base. Here is a prototype https://twitter.com/andrewprojdent/...