Simple answer is yes and no. Yes it will increase speed of some function
related to image reading (AFAIR only for tiff) and processing, but not for
OCR (the slowest part of whole process). And you will face new issues.
Search forums (user and dev) and issue tracker for more details.
Zdenko
po 2
I found the unicode for Baybayin (which is the language) which is this
As for typing out those symbols, there are only a few keyboards that
support it because there is not enough research on the language as of now.
--
You received this message because you are subscribed to the Google Groups
"te
If you know you won't have numbers, what worked for me is blacklisting
numbers. Otherwise you will have to improve the image quality (like
resizing to bigger size and sharping the edges)
On Mon, 29 Apr 2019 at 12:01, fady taher wrote:
> the model keeps outputting (5) instead of (S), I tried to g
(base) compute@instance-1:~$ ls -l /usr/include/CL
total 840
-rw-r--r-- 1 root root 75474 Feb 21 2018 cl.h
-rw-r--r-- 1 root root 307101 Jan 25 2018 cl.hpp
-rw-r--r-- 1 root root 304701 Jan 25 2018 cl2.hpp
-rw-r--r-- 1 root root 5122 Feb 21 2018 cl_d3d10.h
-rw-r--r-- 1 root root 5116 Feb
Hi,
inverting the image gives the correct results. Also cropping the image just
around the text works.
Lorenzo
Il giorno lun 29 apr 2019 alle ore 19:11 Jason ha
scritto:
> Apologies for such a simple question but this is a super simple test case
> and I don't understand why it isn't working. T
Would it not increase inference speed? I'm looking for anything that may
help with that
--
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.
Very nice! You make it look easy.
Maybe I can use it to see why Qt mouse positions are different to what
tesseract reads.
Thanks!
On Monday, April 29, 2019 at 8:57:03 PM UTC+2, zdenop wrote:
>
> here is code
>
> Zdenko
>
>
> po 29. 4. 2019 o 20:44 Zdenko Podobny >
> napísal(a):
>
>> I did with p
I was reading the docs (
https://tesseract-ocr.github.io/4.0.0/a02186.html#a96899e8e5358d96752ab1cfc3bc09f3e
)
and came across this apparent conflict and also noticed that the two
paragraphs have overlapping content (i.e. datapath, language)
*The datapath must be the name of the parent dire
here is code
Zdenko
po 29. 4. 2019 o 20:44 Zdenko Podobny napísal(a):
> I did with python ;-)
>
>
> Zdenko
>
>
> po 29. 4. 2019 o 20:11 Philcat napísal(a):
>
>> By the way, what application are you using to visualize the UZN?
>> Thanks
>>
>> On Sunday, April 28, 2019 at 2:24:41 PM UTC+2, zden
Thank you for looking into this and confirming I am not crazy.
On Monday, April 29, 2019 at 1:11:40 PM UTC-4, Jason wrote:
>
> Apologies for such a simple question but this is a super simple test case
> and I don't understand why it isn't working. This simple image contains the
> words "PASS" a
I did with python ;-)
Zdenko
po 29. 4. 2019 o 20:11 Philcat napísal(a):
> By the way, what application are you using to visualize the UZN?
> Thanks
>
> On Sunday, April 28, 2019 at 2:24:41 PM UTC+2, zdenop wrote:
>>
>> Which mean if you create correct uzn file you will get what you need...
>>
Please provide output of:
ls -l usr/include/CL
+ attach (compressed) config.log
BTW: if do not plan to contribute to OpenCL part of tesseract do not waste
time with its compilation. It does not bring you expected speed increase +
there are long term not fixed issues (search issue tracked )
Zde
Also tried "CPPFLAGS+=-I/usr/include/CL/ ./configure --enable-opencl"
didn't work either
On Monday, 29 April 2019 18:15:58 UTC+3, C493d wrote:
>
> I get this while running ./configure with the OpenCL option (checking for
> clGetPlatformIDs in -lOpenCL... no)
> Any ideas how to fix this?
>
--
Y
I tried "export OPENCL_INC=/usr/include" and "export
AM_CPPFLAGS=/usr/include/CL",
also tried changing inside configure.ac before running ./autogen.sh but no
luck
On Monday, 29 April 2019 18:48:28 UTC+3, zdenop wrote:
>
> Than try to set variable OPENCL_INC to opencl header directory (above
> C
By the way, what application are you using to visualize the UZN?
Thanks
On Sunday, April 28, 2019 at 2:24:41 PM UTC+2, zdenop wrote:
>
> Which mean if you create correct uzn file you will get what you need...
> Zdenko
>
>
> ne 28. 4. 2019 o 14:19 Zdenko Podobny >
> napísal(a):
>
>> Your uzn file
ubuntu@tesseract-ocr:~/TEST$ tesseract fixed.png - --psm 6 --dpi 300
--tessdata-dir ~/tessdata_fast
PASS wee
ubuntu@tesseract-ocr:~/TEST$ tesseract fixed.png - --psm 6 --dpi 300
--tessdata-dir ~/tessdata_best
PASS AYE
ubuntu@tesseract-ocr:~/TEST$ tesseract fixed.png - --psm 6 --dpi 300
--tessdata-d
Apologies for such a simple question but this is a super simple test case
and I don't understand why it isn't working. This simple image contains the
words "PASS" and "FAIL". "PASS" is recognized but "FAIL" comes out as
"wee". What can I do to get it to detect "FAIL" properly?
I'm using the dem
Than try to set variable OPENCL_INC to opencl header directory (above
CL/cl.h or OpenCL/cl.h)
If it does not help try to set opencl header directory to AM_CPPFLAGS.
Zdenko
po 29. 4. 2019 o 17:24 C493d napísal(a):
> I have the headers installed and the OpenCL files that come with CUDA
>
> On M
I have the headers installed and the OpenCL files that come with CUDA
On Monday, 29 April 2019 18:17:00 UTC+3, zdenop wrote:
>
> Install OpenCL?
>
> Zdenko
>
>
> po 29. 4. 2019 o 17:15 C493d >
> napísal(a):
>
>> I get this while running ./configure with the OpenCL option (checking for
>> clGetPl
Install OpenCL?
Zdenko
po 29. 4. 2019 o 17:15 C493d napísal(a):
> I get this while running ./configure with the OpenCL option (checking for
> clGetPlatformIDs in -lOpenCL... no)
> Any ideas how to fix this?
>
> --
> You received this message because you are subscribed to the Google Groups
> "t
I get this while running ./configure with the OpenCL option (checking for
clGetPlatformIDs in -lOpenCL... no)
Any ideas how to fix this?
--
You received this message because you are subscribed to the Google Groups
"tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails f
UPDATES:
It works fine now. All I did is git clone the latest master branch and
build from source. Then the user-pattern and configs can work.
On Monday, April 29, 2019 at 12:37:46 PM UTC+8, phillip...@gmail.com wrote:
>
> UPDATES:
>
> I trained a file and put it in tessdata. The output is bette
the model keeps outputting (5) instead of (S), I tried to go with finetune,
but it seems the process messed up the whole model ... how can I increase
the model accuracy
--
You received this message because you are subscribed to the Google Groups
"tesseract-ocr" group.
To unsubscribe from this
Tesseract generates unicode output after recognizing.
Are there any unicode points for symbols that you have used?
How do you type out those symbols?
On Mon, 29 Apr 2019, 13:42 anne, wrote:
> I used this line
>
> tesseract [lang].[fontname].exp[num].tif [lang].[fontname].exp[num]
> batch.noch
I used this line
tesseract [lang].[fontname].exp[num].tif [lang].[fontname].exp[num]
batch.nochop makebox
from
https://github.com/tesseract-ocr/tesseract/wiki/Training-Tesseract-%E2%80%93-Make-Box-Files
--
You received this message because you are subscribed to the Google Groups
"tesseract-
I used this line
tesseract [lang].[fontname].exp[num].tif [lang].[fontname].exp[num]
batch.nochop makebox
from
https://github.com/tesseract-ocr/tesseract/wiki/Training-Tesseract-%E2%80%93-Make-Box-Files
On Monday, April 29, 2019 at 11:51:18 AM UTC+8, anne wrote:
>
> Haloo, I want to train Tes
26 matches
Mail list logo
