Hi Tom,

the compression artifacts are of course easy to avoid, but the “ghosting” 
in the image is definitely a severe problem. I noticed that, too, but I 
have no idea what the reason could be. Again, a different AD converter 
could help. I already tried to clean the “ghosting”, but had no success.

Another interesting fact, I found a different OCR solution called 
"https://ocr.space/";, which seems to handle my kind of images pretty well. 
It's not a free service (decent free tier, tough), but that could be an 
alternative solution, if I can't manage to get a better image quality.

I would still prefer a local solution with tesseract, will post my updates.

Greetings,
Chris

On Sunday, November 13, 2022 at 9:12:57 PM UTC+1 tfmo...@gmail.com wrote:

> The image has "mosquito noise" around the characters which indicates that 
> it's been compressed with JPEG or similar algorithm. You should definitely 
> try to avoid any compression at this low a resolution.
>
> I think your idea of investigating different video capture devices is a 
> good one. It looks to me like there is horizontal "ringing" or an echo in 
> the signal which is showing up as two ghost images slightly offset from 
> each other in the X axis. With a clean signal you'd have a much easier 
> time. If you are forced to deal with this, you can construct your filter 
> matrices to operate in the X axis, but leave the Y axis untouched.
>
> Tom
>
> On Saturday, November 12, 2022 at 12:57:55 PM UTC-5 goaf...@gmail.com 
> wrote:
>
>> Hi,
>>
>> I want to OCR this kind of image, which is from a video grabber, 
>> unfortunately of pretty bad quality. With the default options of tesseract, 
>> it's pretty useless.
>> Before I start digging deeper into training tesseract, I would love to 
>> hear some recommendations. Would it be possible to achieve a good result 
>> from this kind of image with proper training?
>> Any further ideas/tips would be appreciated!
>>
>> Greetings,
>> Chris
>>
>> [image: temp2.jpg]
>>
>>

-- 
You received this message because you are subscribed to the Google Groups 
"tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tesseract-ocr+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tesseract-ocr/1918b0ba-db56-4fd9-af3e-7079445757c0n%40googlegroups.com.

Reply via email to