How I got into FOSS

February 24, 2009 at 2:55 pm (Uncategorized)

I remember a time when I was in my 2nd year and someone I now really hate (for personal reasons) told me how great Linux is. At that time I replied saying “MS-DOS is just as powerful as Linux”.

That was then, and this is now.

I currently contribute to the Indic OCR project. It all began in my 2nd year, when a group of professors from ISI Kolkata visited our college and delivered a seminar on Digital Image Processing. OCR was one topic that was discussed at length. I wanted to work on OCR but never found the resources since I had no computer at that time.

After around a year of active involvement with our LUG and having built CodeCracker, it was summer internship time. I applied to the Google Summer of Code for a project called LTSP (Linux terminal server project), and was assigned a mentor, which is pretty close to being selected. My application was ultimately rejected and I was dejected.

I never gave up my love for OCR and applied to a few research institutes’ fellowship schemes. My project proposals were nothing fancy. I simply wanted to implement existing algorithms for Indic OCR and then free the code, but those proposals were rejected.

I went back home in the summer vacations pretty upset. It had been a month into the summers and I had no internship. Shreyank and I had applied for NOSIP , but I did not like the projects that NOSIP had. I then tried out GNOME hacking. I read a few online tutorials and used jhbuild and had some fune with gtk.

It was then that I applied for an internship at Red Hat. I got a pleasant reply from a person called Mr. Sankarshan, which was pretty unexpected. After rejecting 3 of my proposals, he finally agreed to mentor me on my OCR proposal.

I had managed to go through the Tesseract-OCR source code, which is a pretty good OCR engine for english, and figured out how to alter the images being OCRed via C++ routines. I soon figured out what I have to do to accommodate Indian languages like Hindi and Bengali. I played around with the scanned image, which is essentially a matrix of 1s and 0s and devised my own algorithms. I wrote code that could clip the “maatraas” between two successive “aksharas” and then use the existing recogniser to OCR Hindi/ Bengali text-images. Via the internships I also learned how to plan my work and how to report it properly.

It turned out that Indic OCR is a very important project for many people around the world, and an organisation named Sarai decided to award me a fellowship. So here I am, working on making the first freely available Indic OCR system.

My only advice to beginners: Do it only if playing with code thrills you, otherwise find something else that does.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: