Some test text!

menu
Document Resolutionkeyboard_arrow_down

Choosing the Right Resolution for OCR in Python

By definition 1 point = 1/72 of an inch, and a resolution of 72 dots per inch on paper (or pixels per inch in the context of digital media) implies that 1 dot (or 1 pixel) = 1 point.

Roughly speaking, to have a 12 point font means that glyphs are bounded by boxes of 12 points in height maximum (as in older days they would have been cut in 12 point tall blocks of wood), so that at 300 PPI 12-point-font glyphs would be inscribed within 50x50 pixel bounding boxes.

A good rule of thumb is that more pixels in a glyph is better than less, OCR quality starts to diminish exponentially if glyph bounding boxes get smaller than 20x20 pixel.

Get the answers you need: Support

close

Free Trial

Get unlimited trial usage of PDFTron SDK to bring accurate, reliable, and fast document processing capabilities to any application or workflow.

Select a platform to get started with your free trial.

Unlimited usage. No email address required.