Convert NDPI to HOCR

Turn NDPI images into HOCR. An overview of the whole slide is read, at most 4096 pixels on its longer side; a smaller slide keeps its own size.

Converted in an isolated sandbox on our serversDeleted after 24 hours

Drop NDPI files on the dialor turn it to pick what to do

Text with word positions from a slide.An hOCR file that records where each word sits on the page.

  1. What you get

    One hOCR file. It holds the text that OCR reads from the slide as HTML, with the position of every word on the page.

    NDPI files are whole-slide microscopy images from Hamamatsu scanners. hOCR is OCR output, so the result is text with positions, not an image.

  2. What to know about size

    The full-resolution pyramid of an NDPI file is far too large for one picture. Keep this in mind when you expect text from a whole slide.

  3. Setting the language

    Set Language to the languages the text is in, so OCR knows which letters to expect. Naming the right language matters most for Korean, Japanese and Chinese.

    Separate several languages with commas, such as en,ko.

NDPI and HOCR.What each format is, and the options this conversion takes.

  1. 01 · language

    Language

    The languages the text is in, so OCR knows which letters to expect. Without it the text is read as English, so letters such as é, ß and ł come out wrong and other scripts, such as Cyrillic, Arabic, Thai, Korean or Chinese, are not read. Separate several with commas, such as fr,en.

The formats

From · NDPI

Hamamatsu NDPI

Whole-slide microscopy images from Hamamatsu scanners.

Extensions
.ndpi
MIME type
image/x-ndpi
Kind
Images
To · HOCR

hOCR

OCR output as HTML with the position of every word on the page.

Extensions
.hocr
MIME type
text/x-hocr
Kind
Text and data

In your code.The same NDPI to HOCR conversion through the API, for your app or your agent.

terminal
$ cv convert photo.ndpi photo.hocr
$ cv convert photo.ndpi photo.hocr

Questions.What people ask about NDPI to HOCR.

01OF 04

It contains the text OCR finds, as HTML, with the position of every word on the page.

Set Language to a list separated by commas, such as en,ko.

Naming the right language matters most for Korean, Japanese and Chinese. Set it for any text so OCR knows which letters to expect.

You get one file.

More conversions.Other ways in and out of NDPI and HOCR.