It highly depends on quality of the image. For clean, high-res images TextRecognize should work just fine. With some of the new additions, you would also be able to get the location of each chunk of text to perform layout analysis, etc.
However, quality of PDFs may vary a lot and I'm sure there will be cases that TextRecognize is not capable of recognizing text. Sometimes upsampling the image would help. For example, call
ImageResize[image,Scaled[2]]
If you give us a sample of your PDFs, we can send a more specific suggestion.