Currently the OCR engine will process all PDF documents, both image-only and searchable. There is an option to exlude all PDFs from the OCR process which is configurable in the OCR Format, but it also would skip both searchable and image-only type PDFs.