Add ocr'ed text back to image and generate a PDF

[ramiromagno]

It would be great if this package supported adding back the retrieved text from a raster to PDF format.

For example, using tesseract directly from the command line makes this possible in one single command:

tesseract --dpi 600 --oem 2 input_01.png output_01 pdf

[brshallo]

Second this.

Relatedly, would be nice if could take in a .pdf that contains some images and convert these to editable text, returning a new .pdf like how done with adobe.