I was looking for a way to measure how dirty is the document in image processing. Example images below:

a. Dirty Backgrounds

enter image description here

b. Dirty Backgrounds with Text

enter image description here

Can someone give me an initial idea to start my analysis?

  • $\begingroup$ Background only, or full document? The question sounds difficult. What do you want to use it for? $\endgroup$ – Laurent Duval Mar 5 at 23:16
  • $\begingroup$ @LaurentDuval The background and the full document. I just want to assess how dirty the document is. $\endgroup$ – alyssaeliyah Mar 6 at 5:20
  • 1
    $\begingroup$ This will require you to spend some time thinking about what exactly you associate with the colloquial term "dirty". It can mean a lot of things. For instance, I guess we're not talking about dust or dirt but maybe more about stains and spills. But what does that mean, what is a stain? Patch of different color? What sizes are relevant, what relative color change? General inhomogenity of the background? That could be measured. What else? Cracks/Folds? I suggest you make a list of possible "contaminants" and their severity and how they would materialize in the image. $\endgroup$ – Florian Mar 6 at 9:47
  • $\begingroup$ Indeed, the "just" seems to imply a form of simplicity. If there is not baseline method available, it's a good sign that it could be complicated, without more details, assumptions, modelling etc. $\endgroup$ – Laurent Duval Mar 6 at 12:02
  • $\begingroup$ @Florian -> Is it ok to measure the amount of non-white pixels, then based from that give the document image quality score? $\endgroup$ – alyssaeliyah Mar 9 at 16:48

Your Answer

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

Browse other questions tagged or ask your own question.