Why error in Tesseract text recognition?

Question

Python

Hello everyone, what is the error in text recognition? No matter how hard I try to apply various filters, converting to RGB, HSV, the text is not recognized correctly for me.

Actually, here is the finished image, and also threw off the second option. As for me, green is better :)



And here is the result when the text has already been read and displayed in the text in the console:

As you can already see, there are a couple of unnecessary letters .... Everything is quite simple in the code. I take a screenshot in the right place on the screen, after which I convert the image to HSV, namely, I make the text green and the background black. How to solve it? I also saw on some forum that if the final version of the picture is converted to RGB, then there are more chances for the picture to be recognized, but it didn’t help me personally :C
config = r'--oem 3 --psm 13'