Se a página contiver texto que o OCR do Acrobat possa entender, talvez você consiga usar o recurso "Reconhecer texto", pois aprendi deste link .
A partir do último post dessa página:
In Acrobat X I can use Tools/Recognize Text/ Aa In This File which will deskew and process page(s) for OCR, i.e. make it a searchable document. When I do this on a multiple manually scanned document which has pages upside down, the deskew rotates them 180 degrees, that is corrects the manual scan orientation.