- Published on
Namecard OCR in Javascript
- Authors
- Name
- Martin Andrews
- @mdda123
One thing that has always bothered me as a regular MeetUp participant is that I usually arrive home with a small collection of business cards (aka namecards in Asia), and the most direct method of getting that data into my address book was manual typing. Therefore, doing OCR via a mobile camera's phone seemed like an interesting project that would solve a problem that I already had (as evidenced by the dozen neat stacks of cards waiting to be typed in on a table at home).
I decided to prototype in Javascript (rather than Python, or more amitiously Scala) because various component pieces already had nice interfaces to nodejs, which could also run as the backend web-server. In addition, the actual code (run initially on the server) would be able to run on the phone itself.
Presentation Link
I recently gave a presentation about this project at the Singapore Talk.js MeetUp Group.
![Presentation Screenshot](/_next/image?url=%2Fstatic%2Fimages%2Fblog%2F2014-08-25_OCR-v0.01_600x390.png&w=1200&q=75)
If there are any questions about the presentation, please ask below, or via the MeetUp group (or at the next NodeSchool!).
![Presentation Content Example](/_next/image?url=%2Fstatic%2Fimages%2Fblog%2F2014-08-25_OCR-v0.01_6_600x390.png&w=1200&q=75)