Separate images of white text and black text.

Question

1.00/5 (1 vote)

See more:

, +

C#

Hello everyone,
I am a student currently working in a image processing project.
My task is to separate white text and black text images.

<a href="https://drive.google.com/open?id=0B81c7NmWqSF9UGFYTW0wRUNQM2c">3.jpg - Google Drive</a>[<a href="https://drive.google.com/open?id=0B81c7NmWqSF9UGFYTW0wRUNQM2c" target="_blank" title="New Window">^</a>]

<a href="https://drive.google.com/open?id=0B81c7NmWqSF9eTFBQ0pYWGRfeUE">5.jpg - Google Drive</a>[<a href="https://drive.google.com/open?id=0B81c7NmWqSF9eTFBQ0pYWGRfeUE" target="_blank" title="New Window">^</a>] 

Can anyone suggest method/ parameter to differentiate these images.

Thanks in advance.

What I have tried:

Tried various filters but unable to differentiate as background is varying in images.

Posted 27-Jun-16 5:55am

Member 9379249

Updated 27-Jun-16 6:33am

Add a Solution

Comments

Sergey Alexandrovich Kryukov 27-Jun-16 12:41pm

Please don't re-post. Instead, continue the discussion on the page of your original question.
—SA

Member 9379249 27-Jun-16 12:46pm

For me both problems are different. I want separate binarization flow for black and white text images, while low contrast images need to handle separately. I have many images of metallic surfaces, those have various noises/variables contrast. All noises can not be handled with single transformation/filter, I think.

Sergey Alexandrovich Kryukov 27-Jun-16 13:48pm

They are not different enough to make separate question posts. As I said before, you hardly need binarization. Splitting the discussion into two different pages is not helpful at all; and it would be very helpful to keep it all on the same page. You did not accept my previous answer and already asking a separate one. That was me who pointed out that image filtering is more relevant to your problem than image classification, and I agreed that both approaches might be needed.

I clearly understand added complexity of your problem, but I'm sure it's solvable with some acceptable fidelity.
Please see my answer on this page. This is a really good idea which helped me a lot.

—SA

1 solution

Add a Solution

Add your solution here

Treat my content as plain text, not as HTML

Preview 0

…

Existing Members

Sign in to your account

...or Join us

Download, Vote, Comment, Publish.

Your Email
Password
Forgot your password?

Your Email
This email is in use. Do you need your password?
Optional Password

I have read and agree to the Terms of Service and Privacy Policy
Please subscribe me to the CodeProject newsletters

When answering a question please:

Read the question carefully.
Understand that English isn't everyone's first language so be lenient of bad spelling and grammar.
If a question is poorly phrased then either ask for clarification, ignore it, or edit the question and fix the problem. Insults are not welcome.
Don't tell someone to read the manual. Chances are they have and don't get it. Provide an answer or move on to the next question.

Let's work to help developers, not make them feel stupid.

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

Sergey Alexandrovich Kryukov · Answer 1 · 2016-06-27T06:33:00

You just generalized your previous question. But my answer was already general enough: Differentiating low and high contrast text images.

I cannot tell you want exact image transformations/filters you have to apply, because it depends on the platform and imagine libraries you are using. But each and every non-nonsense image processing of vision library has the methods for contrast manipulations, as well as dealing with noise. Also, you may need to convert color images to gray-scale (converting to black and white is hardly needed, but it depends on the library, available implementations of blob recognition, OCR, and the like).

I already advised you to create an experimental application where you can play with different transformations/filters. It is really important, to get some experience, and to figure out interactively what really works. This approach is really productive.

—SA