Problem with ocr

[Frenzie]

Originally Posted by ichnilatis

You mean I can make the file with Notepad with just the above mentioned text for ocr?

One more question: Why there are only two options for the text language? What should be the second option instead of "Chinese"? Each user has to make the change manually in the defaults.lua?

That's correct, while you could copy the whole file it's much clearer if you only put in the lines you've changed.

With regard to the OCR, it's not the most polished feature. An ereader is not a very powerful device, so we only feed the OCR engine very tiny parts of the page. It can be nice for the occasional use, but if you know in advance it might be better to do the OCR on your PC first. It'll do it a heck of a lot faster.

A lack of spaces in Greek certainly sounds odd. We'll have to check what's up with that. Not detecting spaces properly can sometimes happen in the Latin alphabet with italics, so I didn't think much of it.

[ichnilatis]

I finally upload the screenshots I mentioned.

I still can't understand why there are only two options for the text language? What should be the second option instead of "Chinese"? Each user has to make the change manually in the defaults.lua?

You're right, Frenzie, about ocr in ereaders and I understand this weakness. I just mean to use this feature for occasional use. Also it's very useful just to highlight an unrecognized text.

[Frenzie]

As I said, it's a lack of polish for that particular feature. There isn't really a reason other than that the person who implemented it isn't an active contributor anymore.

[ichnilatis]

Originally Posted by Frenzie

As I said, it's a lack of polish for that particular feature. There isn't really a reason other than that the person who implemented it isn't an active contributor anymore.

Ο.Κ. Thank you very much for your help and your patience with me.

[ichnilatis]

Originally Posted by Frenzie

Something like that, yes. If you want to keep it, make sure to put it in persistent.defaults.lua.

Dear Frenzie,
I created the persistent.defaults.lua file by copying only the ocr lines from defaults.lua. However, when I updated KOReader, these settings did not apply as defaults.lua was replaced. How can I maintain persistent.defaults.lua settings after each update? Do I have to copy-paste these lines into default.lua every time I update KOReader?

[pazos]

Originally Posted by ichnilatis

Dear Frenzie,
I created the persistent.defaults.lua file by copying only the ocr lines from defaults.lua. However, when I updated KOReader, these settings did not apply as defaults.lua was replaced. How can I maintain persistent.defaults.lua settings after each update? Do I have to copy-paste these lines into default.lua every time I update KOReader?

The file should be named "defaults.persistent.lua", not "persistent.defaults.lua".

If you're using windows as your OS do the following:

1. with Windows explorer copy "defaults.lua" and paste it as "defaults.persistent.lua"
2. open the new file with Notepad++
3. Remove all the lines you won't plan to override (most of them I guess)
4. Edit the rest of lines to match your preferences

That's it. Changes is defaults.persistent.lua should persist, as the name implies

If for some reasons you won't see your changes in the program that means you made a typo. The program will prevent you to load "broken code" and will fallback to default.

Here is an example of a typo:

SOME_STRING = "Unfinished string
SOME_OTHER_STRING = "OK string"

Please keep in mind that the file/logic was developed long time ago when KOReader was used mostly by hackers. It won't prevent you from doing stupid things, like asigning a number when the program expects a string. The program will use these changes even if that means it will crash it later on.

[ichnilatis]

Thank you a lot! I will check it in the next update.
At least, it's not too much to copy-paste it after every update...

[estral]

I have now problems with ocr. What can I do?

[Highroller]

I've had my share of OCR troubles too, especially when dealing with documents for work. It can be really frustrating when the technology doesn't seem to catch up with the languages or details we need it to recognize. What really turned things around for me was finding a solution that could handle a wide range of languages and document types accurately. That's where Identity Verification services, like those offered by ID Analyzer, come into play. Their advanced OCR technology is designed to accurately extract data from a variety of documents, even those that are out of focus or have low resolution. This capability is especially useful in a professional setting where accurate data capture and multilingual support are critical. Their system has proven to be reliable for me, offering peace of mind that the information is being processed correctly.