Issue with binarization of gray colored text of Tesseract

Question

I am using Tesseract to recognize some values from the photo which is taken via camera of a phone. My problem is that I cannot recognize numbers which are written in gray color. I know that tesseract performs binarization on the input image. I manually performed the binarization and I realized that all gray texts disappear after binarization. As you can see in the binarized image, there are not any values appeared infront of sollwert and pos/neg toleranz. I can manually set a high threshold to keep my gray colored text,but I need a automatic way in my app. Beside that, the high threshold returns a mostly black image which it also lost some of information. Do you have any suggestions on how to keep my gray color text after binarization? Thanks in advance