Could anyone please teach me how to clean up the background of a scanned document, using PS?
Does PS have text recognition function that I can use to extract the text, because essentially the text is all I want?
Thanks!
5
-
Hi user22439, I do not think this image is salvageable using adjustments from Photoshop. You will need to use different software called Optical Character Recognition to scan the document for text. I have used some online ones like onlineocr.net.– AndrewHCommented Oct 15, 2015 at 16:25
-
Hi, thanks for the reply. I just tried the website you linked... it gave me this "No recognized text !"– user22439Commented Oct 15, 2015 at 16:29
-
2I am not surprised, I don't think you can salvage this image. I would advise you take another try at scanning the text again or re-writing the text yourself. The reason i believe the image cannot be saved is all of the black lines are the same color as the text which makes it probably impossible to just see the text.– AndrewHCommented Oct 15, 2015 at 16:33
-
Maybe instead of black and white scan a grayscale image then we can do something. Oh and by the way by scaling down i can make it grayscale, adjust and do a successful char recognition for 70% on it. ;) Scan again.– joojaaCommented Oct 15, 2015 at 20:01
-
2Retype it. Problem solved.– ScottCommented Oct 16, 2015 at 7:26
Add a comment
|
1 Answer
That's a very bad scan! Even software that have OCR might not be able to decode this because of the lack of contrast.
It looks like you use a setting on your scanner to convert it to black and white only; maybe scan in colors or grayscale instead. If you have a gray background and black texts, it will be way easier to make one black and the other white. Right now, it's all pure black and white and it's very difficult to separate the texts from the background.
If this doesn't work, maybe try to photocopy your pages by adjusting the contrast on the photocopier and then scan these photocopies. Sometimes it helps to get rid of the background noise. It's an extra step but at least you might be able to get a clearer scanned image. That's a good trick if you don't want to spend too much time adjusting the levels in Photoshop, or at all. Photocopiers can be adjusted to act as the threshold or curve functionalities in photo editing software.
Then you can use a OCR feature in Adobe Acrobat Pro or any software that has it, but know that the text will still need some editing because it doesn't always give perfect results. OCR will transform your scanned texts as real editable texts.
In Photoshop, you can always adjust the contrast of the text and background by using the "adjustment levels" to make the background white and the text black (see trick here, similar technique: how to compensate 50% opacity white over photograph)
Similar tricks from Smithsonian Institution Archives:
