Pytesseract not detecting periods: '.' properly. Improving recognition?

Viewed 143

This is my first time messing around with image parsing, so I'm not sure as to what I can do to improve this.

Here's the image: enter image description here

(It's from a mobile game)

and here is the list that image_to_string().split() returns: ['Connected', 'to', 'hAMSO4', '69.113194', 'has', 'been', 'copied', 'to', 'the', 'clipboald!']

My first thought was to edit the image (to increase the size of the period) using PIL or something prior to parsing it using pytesseract, but that seems pretty complicated.

I was wondering whether there is a way to provide it with the font that the text will be in to improve the performance?

I apologize for being abstract, but I'm not quite sure how to proceed

0 Answers
Related