OCR Indexing

Since Tiki20, file galleries can index the contents of files with images uploaded to Tiki, by means of "Optical Character Recognition" (OCR), and take the result to feed the search index also.

Tiki relies on https://github.com/tesseract-ocr/tesseract so you need to install as per https://tesseract-ocr.github.io/tessdoc/Installation.html

If you are using WikiSuite, Tesseract is installed by default: https://wikisuite.org/Differences-between-Virtualmin-and-WikiSuite

Server Check helps you confirm that Tesseract is working well, and available to Tiki.

Required Preferences

To enable OCR indexing in Tiki, make sure to activate the following preference:
ocr_enable: Enables Tiki to extract and index text from supported file types.

Optional Preferences

You can further customize OCR behavior with these optional settings:
ocr_every_file: If enabled, Tiki will attempt OCR on all supported files, regardless of other criteria.
ocr_file_level: Allows users to override the default OCR language settings on a per-file basis.

Additional Customization

Tiki also offers several advanced customization options:

• Display OCR status per file.
• Set custom paths for the tesseract and pdfimages binaries via the system $PATH.

The file gallery has two view modes: "Finder view" and the default "Page view". The OCR status for files can only be seen in the "Page view" mode of the file galleries File Gallery.

This is where you can switch between File Galleries modes — from the default Page view to Finder view, and vice versa.

Alias names for this page:
OCR | OCRIndexing | Optical Character Recognition