Computer Vision is all about training. You need to create several hundred (maybe thousands) of sample images of what you can try to analyze in order to train Tesseract to detect it. As soon as you think that you are sufficiently prepared, train with new images. Then new images reappear. The more sample data, the better. Once you have an artificial AI, taking a screenshot from the user's login is trivial, pass it to Tesseract for analysis. Be sure to enable the “Help us in learning with this image” feature in your application if you make a mistake, so you can learn in real time from users.
source share