David .'s profile - activity

I read this excellent answer by Mostafa Sataki, where he describes a way to detect whether an image is blurry. In my case, I want to check whether the text on a photo of a piece of paper is readable. There could be several reasons why the text might be unreadable. These include:

1. Too poor resolution (is this burriness?)
2. Too dark
3. Burred image due to shaking/non-focus

My questions is whether the best approach to detect whether the text is readable, is to go with the answer from Mustafa, or whether I should try this from an entirely different approach from the beginning. I imagine for example that OCR software tries to detect the orientation of the text as a first step. Maybe such detection could be used, because if the software cannot accurately detect the orientation, then it must be because the text is unreadable. Does this make sense, or are there other possibilities?