[jira] [Created] (TIKA-1790) Enhancement for extracting text from pdfs

["Randal Moss (JIRA)" <ji...@apache.org>]

Randal Moss created TIKA-1790:
---------------------------------

             Summary: Enhancement for extracting text from pdfs
                 Key: TIKA-1790
                 URL: https://issues.apache.org/jira/browse/TIKA-1790
             Project: Tika
          Issue Type: Improvement
          Components: example, parser
            Reporter: Randal Moss
            Priority: Minor


This enhancement would attempt to extract more text from multicolored background images in PDFs by using adaptive threshold binarization before applying Tesseract for OCR. It also tries to extract text from vector images inside PDFs by first rasterizing them (using Ghostscript) and then applying Tesseract to the flattened images. The final output would be a text file containing all previously extracted text.

I would want to integrate this as a separate library from Tika that is similar to how the [GeoTopicParser|https://wiki.apache.org/tika/GeoTopicParser] is handled.

The code that I have is still a work in progress and can be found [here|https://github.com/RandalMoss/pdf-search].



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)