OCR not able to find text

Asked by mawesome on 2017-12-10

I'm trying to get it to find text that is here: https://cdn.discordapp.com/attachments/228863083584421888/389265484035391488/Screen_Shot_2017-12-09_at_7.53.46_PM.png

but it just returns "--- no text ---". Is there a fix?

Question information

English Edit question
Sikuli Edit question
No assignee Edit question
Last query:
Last reply:
RaiMan (raimund-hocke) said : #1

no fix.

the ocr/text feature is still at the experimental level since 5 years ago.
many problems and oddities. will only get better later in version 2.

in your case:
such grafically pimped fonts will always need special setups and preparations for use with Tesseract (the engine used inside), which is not possible any way with SikuliX 1.1.x.
Best results can be expected with standard fonts in standard GUI elements.

mawesome (mawesome4ever) said : #2

Hmm... is it possible to train the OCR to recognize that type of font?

RaiMan (raimund-hocke) said : #3

Principally yes ;-)

You have to dive into the details of Tesseract (current Sikuli internally has Tesseract 3.02, but still only uses it like a version 2).
You need a separate Tesseract installation on your system.

When you have your relevant data files ready, then you have to integrate them into the the SikuliX environment.
I do not have any experience with that, but faq 2709 might help to get on the road.

If you eventually get the track to a solution: feedback and SikuliX related questions welcome

Can you help with this problem?

Provide an answer of your own, or ask mawesome for more information if necessary.

To post a message you must log in.