Trained data for Tesseract with LetsGoDigital font
With this trained data, you should do a previous image processing to make it work correctly. I've done it through OpenCV:
- First step. Convert the image to Black&White
- Second step. Apply to the image a Gaussian Blur
- Third step. Apply to the image a Threshold filter
After this steps are applied, the recognition will work fine.
EXAMPLE:
Input ->
Output ->
