What about Dilate and Erode in OpenCV ? https://docs.opencv.org/2.4/modules/imgproc/doc/filtering.html#dilate
I mention my experiments here on the Wiki (which includes a link about Dilation and Erosion algorithms in general used in lots of image processing software): https://github.com/tesseract-ocr/tesseract/wiki/ImproveQuality#dilation-and-erosion Thad https://www.linkedin.com/in/thadguidry/ On Wed, Jan 29, 2020 at 2:01 PM André Castro <andrelob...@gmail.com> wrote: > Dear, all. > > First, I'd like to thank you for maintaining the Tesseract community > alive. Second I'd like to share some questions about the training process I > am using. > > Following the steps in the tutorial, I was able to create the *box/tiff > pairs *and *lstmf* files with a *ttf* font file. The problem I had was > the recognition was barely adequate for the font provided. I realized the > data I was testing was corrupted with lots of pixels missing after applying > the filters for text segmentation. Is there any way to use OpenCV or any > filter for deteriorating the *tiff* file? I could see the Tesseract > includes extra pixels in the borders of the characters. Is there any > parameter to remove instead of adding? > > Thank you so much! > > *Training tiff File contains:* > > [image: Screenshot from 2020-01-29 16-48-30.png] > > > > > > *Image after processing contains:* > > > > > -- > 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/b1962b86-4963-4020-9182-2d28e78162e6%40googlegroups.com > <https://groups.google.com/d/msgid/tesseract-ocr/b1962b86-4963-4020-9182-2d28e78162e6%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- 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/CAChbWaMvCbMe%3DjgUf1Ho-z_GBcb5J%3DSOV4y_5-FqRuZWSOwnPA%40mail.gmail.com.
