When converting a standard image-based PDF into a searchable PDF, the goal is to make the content accessible to software without disrupting what the human eye sees. In the PDF specification, this is achieved by inserting a content stream containing font resource dictionaries and text operators set to transparent rendering.

Because the text stream is mathematically aligned with the visible scan, selecting a word on your screen highlights the actual scanned image while simultaneously selecting the underlying Unicode characters in the operating system clipboard.