0
$\begingroup$

I have this image below, I want to get rid of the text 'MSAR IDs' that is in the background. I am extracting the numbers from the image with OCR, the numbers that have a white background get extracted as is, however the numbers that have the text behind them don't get extracted properly and are not correct because they all have some part of the background text behind them. Can anyone please tell me how to get rid of the big text in background, any software will do for me.

(edit)PS- I got information that MSAR IDs are intellectual property of a company so to be clear I got this as part of a freelance work, my work is to put the numbers in an excel, so I was thinking of making my work a bit easier hence put up this post

$\endgroup$
  • $\begingroup$ Have you tried contacting MSAR itself to get non-watermarked files? Because what you're want to do is illegal. It is called theft. In this case the theft of intellectual property. I advice people against helping here, since most of our IT jobs depend on the intellectual property protection. Unless OP can prove he has the MSAR approval. $\endgroup$ – Nash Gold Aug 22 at 12:34
  • $\begingroup$ @NashGold - Thanks for the Msar ID explanantion, i never knew about this. I actually got this and a bunch of images as part of a freelance job with an expiry date (The project expires after few days), my work is to put these numbers in an excel sheet. Since the time is limited I thought maybe I can extract numbers and paste them directly in the sheet. I never knew what MSAR id is and i am not planning to forge the images or do something illegal about it. I am just trying to make my work a bit easier so that it completes on time. $\endgroup$ – Abdul Malik Aug 22 at 14:57
1
$\begingroup$

It is all about filtering, you can see there are 4 colors, i.e. C_Font, C_Paper, C_Watermark_Paper, and C_Watermark_Font, the filter is just f(C_Watermark_Paper) = C_Paper and f(C_Watermark_Font) = C_Font.

Of course you can use GIMP and easily Google references, i.e. http://howto.nicubunu.ro/gimp-remove-watermark/

| improve this answer | |
$\endgroup$
  • $\begingroup$ Thank you for your response, appreciate it. I got this solved by using Google API Vision $\endgroup$ – Abdul Malik Aug 26 at 20:54
0
$\begingroup$

There is a good chance that you could just blur the image a tiny tad so that the noisy structure in the watermark disappears (it basically becomes its own grey color). Then maybe your OCR software can read it? It might help converting it into a greyscale image first.

| improve this answer | |
$\endgroup$
  • $\begingroup$ Thank you for your response, appreciate it. I got this resolved by using Google's API vision $\endgroup$ – Abdul Malik Aug 26 at 20:55
0
$\begingroup$

Got this resolved by using Google's API Vision https://cloud.google.com/vision

| improve this answer | |
$\endgroup$
  • $\begingroup$ If your problem is solved, mark the answer that solved your problem (probably your own) as accepted. $\endgroup$ – wychmaster Aug 27 at 5:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.