You don't really understand what do you want. Do you understand, that there are no any texts in a TIFF image? Do you understand, there is no a TIFF image after you read and parser it, there is only a bitmap; the type of it does not matter?
There is no text, there are only pixels of different color, which can or can not be interpreted as text the way human being do. Doing it is a problem of image recognition
, and working with texts is a very specific part of it called Optical Image Recognition
(OCR). Despite the name, the OCR engines usually do not recognize characters, but rather words or phrases using some word dictionary. It looks more complex but much more effective in practice. This is not easy at all. Moreover, at this moment I don't know really good open-source complete and fully functional product which could be easily embedded in your application and would produce good results.
I have no idea how you ended up having text data dissolved in the bitmap and not having it directly, but it looks like you are trying to do something totally wrong.