- Thank you for the response tgflynn and happy new year. I just wanted to understand the architecture of how one is created. Like I want to understand how it works. If somebody points me to a paper or a book, I would be glad to look at it.
I doubt there are any books specifically on the subject of how to build an online OCR service. If you're interested in the underlying technology of OCR or the orthogonal question of the mechanics of building web applications in general you may find something.
The most basic workflow would be to implement a simple web app that allows uploading an image file, save the the file to local disk, run the Tesseract program and generate a new web page that displays the Tesseract textual output to the user.
There is going to be much more CPU utilization per query than for a typical web application so if you want to scale you'll probably need some sort of distributed load balancing architecture with multiple (possibly virtual) servers.
If you don't care about scaling and just want to test something out you'll at least need a hosting provider that allows you to compile and install software. That probably means a virtual server like EC2 or something similar.
I have a Samsung ChromeBook and I tend to prefer it for on the road use to the MacBook Pro. Much lighter and better battery life. With WiFi both public and via smartphone tethering, it is fine for most tasks that I need on the road.
Crouton is fine if you need a full fledged Linux. Using a 32GB SDHC card for extra file space.
- I think NewMe Accelerator will be a better option. Here is their link ~> http://www.newmeaccelerator.com/ . It is like the YC for African American people. There's alot that you'll find.
Paul Judge is also a very prominent and successful entrepreneur in Atlanta. He's had a number of security related businesses. He's currently working as chairman of Pindrop Security and probably doing other things of top of that. Link:http://www.pindropsecurity.com/about-pindrop-security/
