In this text2image, there is an rgument --fontconfig_tempdir which creates
temp folder where fonts.conf gets added.
I checked /tmp/, no other tempfolder is created( font_tmp.0doGBqWc3I)
Has anybody this issue?
Regards,
Samruddhi
On Tuesday, August 31, 2021 at 7:24:46 PM UTC+5:30 Samruddhi Dhak
Tesstrain.sh work for you ?
On Wed, 1 Sep 2021 at 5:09 PM, Samruddhi Dhake wrote:
> In this text2image, there is an rgument --fontconfig_tempdir which creates
> temp folder where fonts.conf gets added.
>
> I checked /tmp/, no other tempfolder is created( font_tmp.0doGBqWc3I)
>
> Has anybody this
No. Tessstrain.sh didn't work. I am running tesstrain.sh on cygwin.
Command->
*$ ./src/training/tesstrain.sh --fonts_dir %WINDIR%/Fonts/ --lang eng
--linedata_only --noextract_font_properties --langdata_dir 'C:/Program
Files/Tesseract-OCR/langdata' --tessdata_dir 'C:/Program
Files/Tesseract-OCR
Okay,
Wait you are working for English language right?
What kind of dataset you used here.
On Wed, 1 Sep 2021 at 5:53 PM, Samruddhi Dhake wrote:
> No. Tessstrain.sh didn't work. I am running tesstrain.sh on cygwin.
> Command->
> *$ ./src/training/tesstrain.sh --fonts_dir %WINDIR%/Fonts/ --lang
Yes, I am working for eng language.
I am using tessdata.(C:\Program Files\Tesseract-OCR\tessdata)
On Wednesday, September 1, 2021 at 5:57:24 PM UTC+5:30 P007 wrote:
> Okay,
>
> Wait you are working for English language right?
> What kind of dataset you used here.
>
> On Wed, 1 Sep 2021 at 5:53 P
I mean working with font only?
Or images??
On Wed, 1 Sep 2021 at 6:09 PM, Samruddhi Dhake wrote:
> Yes, I am working for eng language.
> I am using tessdata.(C:\Program Files\Tesseract-OCR\tessdata)
>
> On Wednesday, September 1, 2021 at 5:57:24 PM UTC+5:30 P007 wrote:
>
>> Okay,
>>
>> Wait you
For images.
I have to create my own trainneddata for my images. So for that I am
following steps mentioned in this
documentation
https://tesseract-ocr.github.io/tessdoc/tess4/TrainingTesseract-4.00.html
As per the steps I have created box file, lstm file and unicharset file.
And next step is to
Check this
https://github.com/tesseract-ocr/tesseract/issues/1685
On Wed, 1 Sep 2021 at 6:22 PM, Samruddhi Dhake wrote:
> For images.
> I have to create my own trainneddata for my images. So for that I am
> following steps mentioned in this documentation
> https://tesseract-ocr.github.io/tessdoc
I checked this link. It has no tesstrain reference.
Tesstrain internally calls text2image.exe.
So if I ran text2image.exe, how will I get trainned data? What are the
further steps to get trainneddata?
On Wednesday, September 1, 2021 at 6:25:13 PM UTC+5:30 P007 wrote:
>
> Check this
> https://g
Hi,
Though I am stuck with tesstrain.sh command, I have one question related to
lstmtraining.exe command.
*training/lstmtraining --debug_interval 100 \ --traineddata
~/tesstutorial/engtrain/eng/eng.traineddata \ --net_spec '[1,36,0,1
Ct3,3,16 Mp3,3 Lfys48 Lfx96 Lrx96 Lfx256 O1c111]' \ --model_o
[image: WhatsApp Image 2021-09-01 at 14.39.54.jpeg]hi, i made license plate
recognition system in linux. I used yolov5 to detect the plate. I used
tesseract to read characters but it doesn't read some letters. I want to
train tesseract for this.
https://michaeljaylissner.com/posts/2012/02/11/ad
11 matches
Mail list logo
