Click here to Skip to main content
15,944,942 members
Please Sign up or sign in to vote.
3.33/5 (3 votes)
See more:
I am doing a thesis work on this topic.I came across a couple of articles of this topic in Codeproject,but my field is little different from them.Those all are developed for English language,but i need to develop it for Bangla language.There are no open source OCR for this language for which i need to start to develop that huge work.I expect some expert's opinion about this topic.I need to know the basic process or steps to develop a handwritten recognizer,how to train computer about my handwritten words and how to recognize them(the computer will print what i have written),also i want to know which article of Codeproject should i follow(because each of them use different algorithms,as i'm new in this topic,i can't understand which one will be good for me as my language is different). I use desktop pc and use windows 7(mention it to sure i am not using tablet or windows 8). Thanks for any advice..
Posted
Comments
joshrduncan2012 23-Oct-12 16:32pm    
We will not do the research for you. I am a fellow grad. student, too. Your research prof expects you to do the work yourself. We will, however, answer any specific questions relating to existing code you might have once you get started coding the OCR project.
Sergey Alexandrovich Kryukov 23-Oct-12 16:51pm    
You are absolutely right. But we don't have to do research for anyone. Still, there is nothing wrong with exchange of open information and idea. After all, openly published works are open, and everyone will be able to find them sooner or later. Original part of work goes beyond reading of literature, in research and development itself (what a strange practice to call the search of literature "research"! real research is developing theory and planning and performing experiments, scientific practice itself, not reading anything).

And I also emphasize importance of original work -- please see my answer.
--SA
ridoy 24-Oct-12 1:42am    
I am not expecting research from anyone,because i know that is my job.But i expect some guidelines from you over this topic for some clarification for mine.I mentioned above that i want to know the basic process or the right way that i can go.Rest of the work will be mine.
sp_suresh 6-Nov-15 1:16am    
Hello ,
Have you got solution for this if yes please reply me thanks as we have to make it as urgent basis.
Thanks in advance

1 solution

A while ago, I found all existing open-source projects, even for printing recognition, very poor. Interestingly, even though some authors are pretty strong in this field and in application mathematics, they are notoriously illiterate in computer sciences and practical programming. In particular, I found that such basic thing as Unicode was a great mystery to them.

However, later on I found much more mature and interesting works on CodeProject. They are old, but it's good for you to know them. At the same time, none of them was ready to use, but main recognition problems were solved pretty well. I looks like dominating approaches were based on neural networks (http://en.wikipedia.org/wiki/Neural_network[^]). Please see:

Neural Network OCR[^],
Creating Optical Character Recognition (OCR) applications using Neural Networks[^],
Neural Network Classifier[^],
A C# Project in Optical Character Recognition (OCR) Using Chain Code[^],
Unicode Optical Character Recognition[^],
Hebb neural network[^],
Backpropagation Artificial Neural Network in C++[^],
OCR Line Detection[^].

Of course, first of all you need your own research, and not even in literature, but in related aspects of applied mathematics, and your own development works, especially in algorithm. The thesis should always be based on original work. I also would advise you (and everyone) to follow my method: start attacking the problem before you read anything at all. Your attempts could be naive, but when you break enough of your teeth on this difficult problem, you will be able to understand theory and work of others much better. Besides, you should not deny a chance to discover something completely original, something which nobody discovered before you.

Wish you the best of luck,
—SA
 
Share this answer
 
Comments
ridoy 24-Oct-12 1:51am    
that's like a great adviser..Thanks SA..+5
Sergey Alexandrovich Kryukov 24-Oct-12 10:15am    
You are welcome.
Are you going to accept the answer formally (green button)?
--SA
ridoy 4-Nov-12 6:49am    
Yes,i accept it,but i have a question to you.Windows 7 has an excellent math input panel and tablet pc input panel.I say it is excellent because it is over 95% accurate and there is also a system for training my computer about my handwritting.Though i don't dream to make such a huge system but you can say i want to create such type of application where i will input some letters of may language and the computer recognize it correctly.So are there any way to know how windows built this system or what are the steps to create such system?
Sergey Alexandrovich Kryukov 4-Nov-12 11:04am    
Apparently, not "windows" but people created the system. Unless we get some source code, all we talk about is just speculations; and we don't have such source code. I must admit the existing Microsoft engine is not bad at all; and, amazingly, it does not require training. Hand-written recognition is fundamentally based on dynamic of the motion of the person who writes. If the process is not recorded, and you only have a bitmap, the most important data is already lost. So, you can collect that digital ink vector information using the Pen PC SDK and then try to interpret it somehow.
--SA
Member 11537261 3-Apr-15 9:59am    
for my application I need the handwriting recognition api or source code in C#.net its very required please give the link where I can get api for my application. I need to extract text from image file of handwriting in editable format so please give me the link where I can get api for my application.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)



CodeProject, 20 Bay Street, 11th Floor Toronto, Ontario, Canada M5J 2N8 +1 (416) 849-8900